El Tiempo en La Vida

07_Mills 2/2/2000 2:55 pm Page 91
Providing Space for Time

The impact of temporality on life course
research
Melinda Mills
ABSTRACT. The article alerts those in the field of quantitative life

course research to the ontological impact of different forms of
temporality. The first section reviews the influence of cosmic
cycles, human development, historical, cultural, social and institu-
tional forms of temporality on life course behaviour. Two central
themes arise. Institutional calendars shape our everyday lives, and
seemingly innocent calendars may influence behaviour. Further-
more, behavioural affects from cosmic and biological time are
increasingly colonized by social constructions of temporality. The
second section outlines how ontological perceptions of time shape
the epistemological approach. The discussion separates temporal
effects from what is an artefact of data, methods and methodology.
Timing and method of data collection, memory, and self-registration
influence results. The author suggests reflexivity, new interpretations
of memory, and blending of methods and sources to improve
research. KEY WORDS • life course • quantitative methods • time
Introduction
The objective of this article is to alert those using quantitative means of

processing life history data to the ontological influence of different forms of
temporality.1 Ontology is the philosophical position that underpins our theoreti-
cal assumptions and ideas about existence. Ontological assumptions frame not
only our understanding of reality, but direct our epistemological approach. An
epistemological standpoint is how we come to gain knowledge from the
external world. Ryder (1965), for instance, advanced the temporal-based argu-
ment that each birth cohort member develops a certain ‘mentality’ due to his or
her immersion in a unique historical context. Armed with this ontological
TIME & SOCIETY copyright © 2000 SAGE (London, Thousand Oaks, CA and New Delhi), VOL.
9(1): 91–127 [0961-463X; 2000/03;9:1;91–127; 012531]
07_Mills 2/2/2000 2:55 pm Page 92
92 TIME & SOCIETY 9(1)
assumption, he then proposed the epistemological approach of cohort analysis.

A quantitative approach to life course research demands the use of temporal-
based census, vital statistics or survey data coupled with statistical methodology
to reconstruct and interpret life histories. By placing ontological consequences
of temporality at the core, three central aspects require disentangling. These
are: what is naturally or biologically determined, what reflects the ontological
reality of historical, social and cultural context of the researcher and respondent
and what is related to data error and the epistemological approach.
The discussion is divided into two sections. The influence of multiple dimen-
sions of temporality, such as natural or cosmic, human development, historical,
cultural, social and institutional forms, on life course behaviour are explored in
section 1. Within this review, two central themes arise. First, that seemingly
innocent temporal categories such as an institutionally contrived calendar may
affect demographic behaviour. The second related area explores whether the
type of temporality that influences everyday life and thus life course construct-
ion has shifted. It is argued that natural and biological forms of temporality are
increasingly colonized by cultural, social and institutional constructions.
Section 2 draws attention to the interplay between temporality and the episte-
mological approach. The ontological construction and experience of temporal-
ity by both respondent and researcher, in addition to selectivity, memory, and
methods of data gathering, influences results. We must furthermore separate
what is a temporal effect from what is an artefact of the data, methods and
methodology.
This exploration of temporality draws from life course research within the
disciplines of sociology and demography. Life course research focuses on how
social processes such as the family, education, employment, health, and migra-
tion domains are structured over the individual life span (Elder, 1974; Hogan,
1981; Mayer and Tuma, 1990). The empirical focus is on the occurrence, timing
in life stage or age, sequencing, duration, or interaction among life events and
domains. Life events are, for instance, marrying, having a child, and entering or
exiting the labour force. Together, events make up life course domains or
careers such as partnership (formerly marital), fertility or labour force. Using
data comprised of the occurrence and timing of life events, researchers then
reconstruct a discernible life path, often referred to as the life course. Life
course research does not only focus on the timing or occurrence of events. There
is also an attempt to understand the way in which individuals shape their life
course, based on a unique and meaningful sequence of events and decisions.
The life course is thus distinguished from the life cycle, which is a predictable
circuit of life stages (Elchardus, 1984). The current life course state or stage of
the individual is a reflection of cumulative past events and an anticipation of
their future life trajectory (Giddens, 1991; Hareven, 1994). Furthermore, when
an individual engages in decision-making or life-planning, she or he often seeks
07_Mills 2/2/2000 2:55 pm Page 93
MILLS : PROVIDING SPACE FOR TIME 93
to ‘colonize the future’ (Giddens, 1991; Nowotny, 1994). However, the indi-
vidual does not make decisions in a vacuum. The construction of the life course
is embedded within historical, cultural, social and institutional (e.g. familial,
religious) contexts. To guide the reader through this discussion, examples are
used from the research area of family formation, which consists of partnership
and fertility behaviour.2
Taking a temporal standpoint to analyse partnership and fertility behaviour in
the life course shifts the focus from static spatial or group-based comparisons of
the life course (e.g. current socio-economic status, residence, education) to
more dynamic analysis of the evolving, natural, constructed and data-related
aspects of time. We are then driven to ask different types of questions. Why are
there twice as many births occurring in December as compared to July in
Bangladesh? Why did yearly Chinese fertility drop so sharply in 1966 and then
rise in 1987? Why do births peak in the spring months and September in
Canada? Why were so many women trying to conceive a child on 9 April 1999?
Why do births climax in January and fall suddenly in December in the former
USSR? Why do one-third of all first marriages in China occur in December and
January? Are seasonal fluctuations in birth and marriage data related only to the
natural rhythms of the cosmos? Why are most life history surveys collected
from persons between the ages of 15 and 49 years? Why do women in India
report their age at first marriage with digits that end with a zero or five or at the
exact age of 18 years? Why are there more missing data for the date of cohabi-
tation as opposed to the date of marriage in some surveys? Do women remem-
ber family events differently from men? How do we methodologically reckon
with the assumption that the present contains pages of the past and expectations
of the future? The answers to these questions are explored through the course of
this study.
Previous life course studies

The recognition of temporality in life course research is not new. Pioneering
researcher Elder argues that ‘any serious study of lives must consider time,
process and context’ (1991: 58). Mayer and Tuma concur that ‘research on the
life course needs to use both a multilevel and multitime framework’ (1990: 7).
This move to rid life course research of static or ‘time-less’ theoretical and
methodological thinking (Blossfeld, 1996: 183) has been embraced by many in
the field (Courgeau and Lelièvre, 1992; De Bruijn, 1999; Elder, 1994; Giele and
Elder, 1998; Hareven, 1994; Ryder, 1992; Willekens, 1990, 1997).
The inclusive reaction to time has most likely occurred for three central
reasons. The first is the realization that the classic multiple levels of the
‘micro–meso–macro’ often only reflected spatial or abstract (non-temporal)
systems. The life course was situated in a context ranging from micro-level
07_Mills 2/2/2000 2:55 pm Page 94
agency (psychological phenomena, action, choice), meso-level interaction

(decision-making in partner or family unit) and macro-level structure (institu-
tions, social structures, world systems). Although embracing multiple levels,
research was unable to embody the more evolving, dynamic and process-based
parallel aspects of time (Willekens, 1990).
This relates to the second catalyst, which is the general shift from static to
dynamic approaches in the social sciences. The classic measurement of time in
quantitative research was discrete. The symbol t was often used to represent the
occurrence of an event or date of observation. Change was then registered by
another discrete state, denoted by time t+1 (Ryder, 1965). As outlined in the
final section of this paper, time is now viewed as dynamic and process-based.
Life events are understood as the outcome of processes, often with a focus
on the duration within a life state (e.g. marriage to first child, conception to
birth). Life course progression, thus, is in a perpetual process of ‘becoming’
(Prigogine, 1980; Young, 1988).
The third impetus is largely a reaction to temporality as narrowly defined
only at one level. In many cases, time is used as a synonym for one aspect of
temporality such as age, cohort, period or point-in-time events (e.g. change in
abortion law) (Mayer and Tuma, 1990). Although these aspects are vital to
uncover the force of temporality, they cannot be synonymous with time or seen
as the only temporal effects. The recognition that time consists of multiple
clocks still appears largely in the work of an innovative handful. In some cases,
the ontological effects are realized, but the empirical approach is insufficient to
encompass all aspects. Conversely, the epistemological approach may lead the
study, automatically excluding additional ontological effects of temporality.
Thus, the rationale of raising awareness of the ontological impact of tempo-
rality in this field is twofold. First, to promote further research of the ontological
force of temporality at multiple levels. This requires going beyond the episte-
mologically driven analyses that often equate temporality with only one or a
few of the clocks of the individual, family, historical, cohort, seasonal cycles,
duration in a process, or point-in-time events. The non-quantifiable aspects of
temporality related to collective social, cultural and institutional experience and
memory remain largely underestimated in such studies. Furthermore, the force
of institutional calendars that differentially schedule everyday life experience
for disparate groups such as men and women, workers and non-workers, are
rarely explicitly addressed. Second, issues of temporality should be introduced
at all phases of the research process. A well developed theoretical approach is
less often coupled with problems related to the temporal interchange with the
collection, processing, analysing and interpretation of life history data by
researchers in this field. This leaves the connection between ontological
temporal assumptions and the epistemological approach relatively unchecked.
Since the use of data organized by clock and calendar conceptions of time is
07_Mills 2/2/2000 2:55 pm Page 95
often central to quantitative life course analysis, the force of theoretical assump-
tions concerning temporality on data and methods is a vital consideration. The
ontological influence of temporality at multiple levels is now discussed in
detail.
1. Different Forms of Temporality
The presence of multiple temporal clocks, or what Nowotny (1992: 424) termed
‘pluritemporalism’, has been a focus of social theorists. As with life course
research, the approach to time within social theory has been shaped by varying
ontological and epistemological approaches. Durkheim focuses on the social
synchronizing and integrative function of time. Marx examines the commodifi-
cation, economic dimension and role of the state in time. Weber looks at the
rationalization of time (Adam, 1990). Mead (1932/1980) argues that the past is
continuously re-created and reformulated as the present unfolds. Adam con-
tends that ‘there is no single time, only a multitude of times which interpenetrate
and permeate our daily lives’ (1995: 12). Schutz and Luckmann (1973) advance
the idea of persistently interacting multiple life-worlds, where the past and
future simultaneously permeate the present rhythms of the human body, planets,
seasons and routines centred upon clocks and calendars. Luhmann (1977/1982)
argues that synchronic (interdependent) and diachronic (dynamic) systems
differentiate themselves from the environment, also adding that the past and
future are mere horizons of the simultaneously occurring present. This relates to
Nowotny’s (1994) conclusion that the future is merely the ‘extended present’.
Likewise, Giddens (1991) posits that by planning our life course we ‘colonize
the future’. We will return to this discussion of the interplay of the past, present
and future at the end of this article. Hägerstrand (1975) offered the concept of
lifelines and recurrent time-space paths that give rise to world lines. Giddens
(1981, 1984), building on Hägerstrand and others, speaks of three temporal
levels. He divided time into experience in daily life, followed by the temporality
of the life cycle and the duration of social institutions produced or reproduced
through regularized practices.
From this brief review, it is clear that choices are required in order to con-
ceptualize and operationalize the ontological influences of temporality on
human behaviour within the life course. On a basic level, we must distinguish
what is a temporal effect from what is attributed to the epistemological
approach. This is addressed within the second section of this article. Within
temporality itself, however, it is useful to clearly separate the natural from what
is symbolically constructed within cultural and social systems. Hägerstrand’s
(1988) conception of time is a useful starting point to understand this division.
Time, he argues, consists of ‘embedded time’, which is tangible and real and
07_Mills 2/2/2000 2:55 pm Page 96
‘symbolic time’, which is an abstract entity to summarize experiences and

observations. Similarly, Domingues (1995) divides social time into ‘identitaire’
and ‘imaginary’ time. Identitaire time is based on the calendar and is repetitive
or cyclical due to its tie with cosmic time. Imaginary time is a free creation of
society, whereby individuals assign symbolic meaning to temporal dimensions
within their lives.
The division between the natural, cosmic and biological or the symbolic,
imagined, socially and culturally manufactured is essential in this article. As the
next section describes, natural or cosmic time (i.e. embedded, identitaire time)
remains relatively fixed across time and space, but temporal constructs, such as
historical, social, cultural and institutional calendars are perpetually in a state of
becoming. They are assigned different meanings and reinvented according to
context. This article stresses the interconnection between different ontological
aspects of time. For instance, the natural seasonal cycles are often mirrored in
agricultural production, religious, educational and sports calendars (e.g.
ice hockey during the winter months). These divisions and interrelations of
temporality are now discussed in detail in relation to family formation in the life
course.
Natural or cosmic time

Natural or cosmic time is based on the relatively fixed, repetitive or cyclical
rhythms of the natural environment, such as the lunar and seasonal cycles. The
ontological consequences of natural or cosmic temporality are often understood
through the epistemological approach of studies of seasonality. These studies
examine the influence of temperature, latitude (the photoperiod) and precipita-
tion, often through statistical (regression) analysis, with demographic behaviour
being the dependent variable or the outcome influenced by these factors.
Several studies have examined how temperature, which is regulated by the
natural cycle of the seasons, influences fertility. Canada, a country that
experiences extremes of heat during the summer and cold during the winter, is
an interesting case. Werschler and Halli (1992) found a statistically significant
association between conceptions and the temperature between 1920 and 1954.
Conceptions were higher, for instance, in years when the temperature in
February was above normal. They also found evidence that intense summer heat
in July reduced coital frequency. Trovato and Odynak (1993) also showed that
Canadian couples timed births to occur in the spring and summer to avoid
extreme winter cold, notably so in 1881. In addition, they noted a peak in births
in September, a finding that is addressed shortly in relation to institutional
calendars.
Due to latitude, individuals also experience different levels of luminosity
during the course of the year. James (1990) linked seasonality of births to
07_Mills 2/2/2000 2:55 pm Page 97
increased sexual activity occurring at different latitudes due to variations in the

photoperiod. Russell et al. (1993) associated seasonal variation in births in
Scotland from 1938 to 1987 with natural factors such as temperature, latitude,
photoperiod and environmental pollution, which they argued influenced sexual
activity.
Finally, fertility has been linked to precipitation or the monsoon cycles,
which are often tied to agricultural production cycles. Bantje (1988) found that
the number of births in Tanzania increased during the dry season of June to
October and declined during the rainy season. This drop in the conception rate
during the hot and wet agricultural season was attributed to physical factors of
exhaustion and stress for female workers. Panter-Brick (1996) also detected that
women’s bimodal birth patterns in Nepal followed a seasonal pattern mirroring
the monsoon. The first conception trough occurred as a result of male out-
migration in the middle of the monsoon period. The second drop was a result of
ovarian suppression and low progesterone levels due to female weight loss
during the arduous agricultural monsoon work period.
Natural temporality may also influence partnerships. Using a sample of an
early English colony in Australia, Gunn (1990) tested whether the inversion of
seasons experienced by moving from England (northern) to Australia (southern
hemisphere) affected marriage patterns. Gunn established that marriages were
regulated not only by the natural environment of seasons, but also by the cycles
of production. The growing and harvesting periods are most prosperous, and
individuals have increased material resources. Marriage behaviour therefore
interacts with and is dependent upon favourable production cycles, which are
often contingent on weather conditions and longer photoperiods.
Although the lunar and seasonal cycles remain fixed over time, some studies
report that natural influences are beginning to subside (Russell et al., 1993;
Werschler and Halli, 1992) with the introduction of air conditioning, central
heating and irrigation. In other words, it is no longer hazardous to have a child
in the cold of winter or unpleasant to conceive a child on a hot summer
day. Within these first examples, it is clear that natural time is not the only
behavioural influence. Some studies have also attributed their findings to the
‘festival effect’, which is discussed later in relation to institutional calendars
(James, 1990; Trovato and Odynak, 1993). Furthermore, the force of the
seasonal cycles works through the medium of the individual’s own biological
system. Changes in the temperature, photoperiod and rainfall influence bio-
logical aspects of hormonal production but also activity levels and psycho-
logical perception of the world. The lack of sexual activity during extreme
periods of heat, for instance, is a blend of biological lethargy and personal
choice or will. Thus, the discussion now turns to the temporal aspects of human
development.
07_Mills 2/2/2000 2:55 pm Page 98
Biological and psychological human development

The next universal aspect of temporality is the internal clock of the individual,
which is divided into biological and psychological development. The onto-
logical effects of human development are often measured via the epistemologi-
cal approach of comparative, descriptive or statistical modeling of life events,
using age (often mean or median age) or life stage as a yardstick of progression
(Neugarten, 1979).
Biological development relates to physiological growth and maturation
through life stages. The biological chronometer is tied to the predictable circuit
of the human body as it moves through its own physical timetable. The clock
begins at birth and continues via maturation, the onset of menopause or sterility,
progressive cell degeneration, ageing, the decline of health, to eventual death. It
is important to note the differences in biological development of men’s
and women’s bodies. The female reproductive clock is often the focus of life
history studies of fertility. Factors such as the onset or end of menstruation,
lactation, post-partum ammenorhea, nutrition or health affect a woman’s likeli-
hood of conceiving a child or being considered a suitable mate (Bongaarts and
Potter, 1983). The ‘biological clock’ is a common expression in discussions
of fertility and parenthood postponement: there is a definitive reproductive
period.
In addition to interrelating with natural cycles, biological time is also increas-
ingly colonized by socially constructed and technologically based perceptions
of time. Young’s (1988) study of ‘chronobiology’ connects biological and
humanly constructed timetables, discussed shortly in relation to institutional
calendars. For example, before the introduction of the contraceptive pill in
the 1960s, the menstrual cycle was often thought of as a regular and cyclical
occurrence tied to natural rhythms. The standardization of menstrual time via
the pill and other contraceptives represents an interaction with socially con-
structed visions of what a regular biological cycle ought to be. This was one of
the first global colonizations of the human body by science and has affected the
natural temporal cycles of women around the world. Building on the work of
Zerubavel (1981), Foster (1996) took a socio-cognitive as opposed to biological
approach to studying the menstrual cycle. Previously, menstrual time had been
viewed as ‘sacred’, often involving physical separation, evil and uncleanliness,
factors that had serious implications for women’s everyday lives. Efforts to
control biological temporal influences (also fertility and mortality) have dis-
rupted the natural cycle to the extent that it is now almost entirely socially con-
structed. Foster concludes, in relation to the menstrual cycle, that the ‘mental
mapping out of what constitutes the elements to this rhythm, so often taken for
granted as a purely natural pattern of time, is a highly social act’ (1996: 544).
Biological development can no longer be understood as wholly fixed or natural.
07_Mills 2/2/2000 2:55 pm Page 99
(Our ability to time and control conception with new reproductive technologies
is touched upon again in relation to the effects of institutional calendars; see
p. 105). Due to the introduction of new technologies, fertility is no longer based
on chance, but is often the result of conscious decision-making by the indi-
vidual, partnership and family. The growing social construction of biological
time is an important observation that will become increasingly relevant as the
force of technology permeates all realms of everyday life.
The individual clock does not only progress biologically. There is psycho-
logical and personality growth related to the life course, a topic that originally
stems from Erikson’s (1950/1963) theory of developmental readiness.
According to Erikson, biological, psychological and social factors determine the
stages of development and sequences that individuals pass through as they
chronologically age. Although people may be of identical age or the same
amount of time may have passed since the last demographic event, they can be
at significantly different levels of biological and psychological development
(Willekens, 1997).
Furthermore, there is interplay between historical-cultural-social contexts
and the perception of psychological development. This is illustrated in
Giddens’s (1991) discussion of self-identity formation in the age of modernity.
One’s ontological outlook in pre-modern and agrarian contexts was shaped by
kinship relations, the local community, religious cosmologies and tradition,
whereas, in the age of modernity, individuals develop their identities in relation
to personal relationships and abstract systems, with a more future-oriented per-
ception connecting the past and present. Abstract systems consist of symbolic
tokens and expert systems, which span time and space (Giddens, 1991).
Symbolic tokens are media of exchange which have a standard value, such as
money. In relation to the life course, symbolic tokens could be a wedding ring,
co-residence or the wearing of black by a widow. Expert systems are those on
which individuals base their knowledge, trust and security: their ‘ontological
security’. Individuals rely on advice from midwives, elders and doctors in
relation to fertility. They may decide to marry instead of cohabiting, due to the
influence of a guardian or a religious figure, for instance, who is viewed as an
expert. Psychological human development and decision-making in the age of
modernity, therefore, become increasingly reflexive and individualistic. An
example will solidify this point.
Let us compare the basic elements of family formation for women in the
largely agrarian area of the northern Indian state of Uttar Pradesh and the
modern western-European nation of The Netherlands. In a 1992–3 survey in
Uttar Pradesh, 63.9 percent of women between the ages of 20 to 24 were
married by the age of 18 (IIPS, 1996).3 A similar 1993 survey in The Nether-
lands found that for women of the same age, a mere 16.9 percent had entered
some kind of consensual or marital partnership. However, only 1.2 percent had
07_Mills 2/2/2000 2:55 pm Page 100
entered into legal marriages not preceded by cohabitation (Latten and De Graaf,
1997: 60). In the first instance, there is a clear disparity in the timing of first
partnership transitions between these two groups of women. As described
shortly, differences such as these are often related in the literature to the more
obvious disparate social, cultural and historical contexts. However, this finding
may also be connected to the ontological effect of the perception of women’s
biological and psychological readiness.
The same survey in Uttar Pradesh shows that over half the girls marry at an
early age, but remain in their parental household after marriage. It is not until
after the gauna ceremony, which signifies the onset of menstruation and entry
into womanhood, that a woman resides with her husband (Mukherjee et al.,
1996). In this way, her authentic entry into a marital partnership is signalled by
her biological readiness to experience the life transition. Although this begins to
change among certain groups of women (higher education, urban), a girl’s
entrance into marriage is still largely regulated by her biological readiness,
ascribed by the family, community, traditional and religious beliefs. Con-
versely, in The Netherlands, there is more emphasis placed on individual
choice, romantic love and preference based on a woman’s own reflexive con-
ception of her psychological readiness. This comparison relates not only to an
individual’s psychological development and socio-cultural context, but also to
the luxury of developing one’s self-identity and planning life events. Russell et
al. (1993), for instance, found that births within wealthier social classes in
Scotland occurred largely during the summer, suggesting that the wealthier had
more resources to plan childbirth.
Although natural and human development clocks are important in governing
the life course, it is too simplistic to see them as the sole regulating factors. As
suggested in the previous example, individual development is interpreted
differently in social, cultural and historical contexts. Historical context is there-
fore my next subject of discussion.
Historical time
Life course researchers (Elder, 1974; Hogan, 1981; Hareven, 1994) have exten-
sively refined the concept of historical time, which is defined by the particular
socio-cultural, economic, physical, political and economic configuration of an
individual’s context (De Bruijn, 1999). This type of time, along with individual
and family or household-based time clocks, is one of the most studied
influences of temporality in life course research. Historical events are seen to
shape a cohort, period or generation. These include war, invasion or sanctions,
famine or surplus, environmental or ecological catastrophes, plague or advances
in health, shifts in ideology such as fundamentalism, capitalism or communism,
and economic depression or boom. The ontological impact of historical period
07_Mills 2/2/2000 2:55 pm Page 101
on the individual or family unit is often traced through the epistemological

approach of age, cohort, or period analysis, discussed in the second section of
this article.
Mayer and Tuma (1990) maintain that a core assumption of life course
research is that ‘life courses emerge in particular historical settings’ (p. 6). Elder
(1974) demonstrated this temporal effect in his book, Children of the Great
Depression. In this book he explores not only the impact of economic change on
individuals, but also the connections between such change and the adult careers
of men and women born during the Depression. Focusing on adaptations made
by the family, Elder notes differences in the timing and sequencing of events in
childhood and later adulthood. Cohorts that had to forgo higher education,
lacked employment or had menial jobs during the Great Depression had less
favourable life chances. Due to these conditions, much of the cohort delayed
marriage and family formation. Conversely, a cohort born during the Second
World War had greater opportunities in an atmosphere of economic prosperity,
had the possibility of military service and later entered employment, marriage
and family formation with greater ease than their predecessors.
Membership in a cohort or period is a central factor in life course oppor-
tunities such as marriage, employment or the ability to begin a family. However,
the temporal category of birth cohort, which is often only arbitrarily constructed
as an epistemological tool by the researcher, may also be a factor related to
value shifts. Foot and Stoffman (1996) studied the baby boom generation that
occurred in Canada in the postwar years of economic prosperity between 1946
and 1966. In 1996, this generation comprised approximately 33 percent of the
population and therefore often set the political, social and cultural agenda of the
country. Foot and Stoffman argued that the current phenomenon of ‘return to
family values’ could be attributed to a cohort effect, rather than a value shift.
This is due to the fact that this large cohort, now between the ages of 30 and 40
years, is in the ‘family’ life stage of mortgages and children. Foot and Stoffman
(1996: 189) predict that at the turn of the millennium, when the large ‘baby-
boom echo’ cohort born from 1980 to 1995 begins to leave the parental home,
values will shift to sex, drugs and new music. ‘Social observers will herald the
arrival of yet another value shift,’ they argue, ‘but the real reason will be
the demographic shift’ (1996: 189). This finding supports Ryder’s (1965: 13)
original theory of the importance of cohort size as ‘likely to leave an imprint on
the cohort as well as on society’. Although one’s spatial or physical location, for
instance, may change as a consequence of migration, persons are born into not
only a particular historical, but also a unique cultural and social context.
07_Mills 2/2/2000 2:55 pm Page 102
Cultural and social time

The cultural and social construction of reality shapes the institutional calendars
that regulate everyday life. Behaviour in everyday life in turn makes up the
social practices that construct the life course and social-cultural structure itself
(Giddens, 1981, 1984). How different cultural and social groups experience
time is discussed in detail in section 2 of this article, which deals with the
empirical challenges for research. The tools used in this approach are often
ethnographic or involve the use of a life history calendar or mnemonic devices
to remember life events (Freedman et al., 1988; Hertrich, 1993).
Cultural time is the subjective conception, use and meaning ascribed to
temporality that is manufactured in different cultures (Hägerstrand, 1988;
Adam, 1995). This discussion, often based in anthropology and geography via
cultural schemas and meaning systems, argues that there are culturally specific
interpretations of time (D’Andrade, 1995). Just as DNA shapes our biological
systems, human culture invents ontology about the nature of the world and
prescribes how individuals should behave in it. Bronowski (1974) originally
suggested that the landscape is human made and shaped by our own symbolic
construction. Hägerstrand (1988: 33) likewise considers time as a ‘wholly
cultural phenomena’. The geographic space of countries or regions, biological
life cycles and the natural cycles of days or seasons are generally constant.
Differentiation stems from how different cultures transform this social space
and time to reflect unique ontological conceptions. By taking this reflexive
position, we attempt to understand different ways of experiencing time. This
places the western researcher beyond the ‘our’ and ‘their’ time distinction so
prevalent in earlier anthropological works (Adam, 1995).
Cultural time may also impact life course events. Salvat et al. (1997)
analyzed 1800–1900 marriage records in the geographically unified Cerdanya
valley of the Pyrenees, an area politically divided by the French–Spanish
border. Their goal was to discover whether sociocultural differences occurred
over time and between French and Spanish communities due to the border. By
comparing seasonality of marriages, they found that the marriage month
progressively differed between French and Spanish, becoming culturally deter-
mined. Differentiation often stems from institutional calendars, such as reli-
gious calendars, discussed shortly.
Social time includes norms, values, rules and responsibilities about the time
at which life events are expected to occur (Adam, 1995; Zerubavel, 1981). Elder
(1994: 6) defines it as ‘the incidence, duration, and sequence of roles, and rele-
vant expectations and beliefs based on age’, which regulate the life course.
Social time has been extensively discussed by Mead (1932/1980), Luhmann
(1977/1982), Elder (1994), Hareven (1994), Nowotny (1994) and Adam (1995).
Elias (1987/1992) suggests that social time, defined as that which extends
07_Mills 2/2/2000 2:55 pm Page 103
former dimensions of space and time to embrace consciousness and experience,

is the result of a cumulative civilization process. As discussed at the outset of
this section, Domingues (1995) further divided Elias’s conception of social time
to include identitaire and imaginary time, or the natural and symbolically
constructed.
Social time is manifested in the area of partnership and fertility behaviour
through calendars that prescribe the ideal age for marriage, having a first child
and ending one’s fertility career. Age is often applied as a standard to judge the
‘proper’ timing of life events, or ‘timeliness’ of life transitions. The individual
is then placed in a logical social temporal order of being ‘on’ time (or not) in
their development (Neugarten and Hagestadt, 1976). Hareven (1994), for
instance, discusses the differences between the importance of the ‘timeliness’ of
life transitions in the USA and Japan. In Japanese there is a specific word,
tekirei, which ‘designates set norms of timing, meaning the age most “fit” for
accomplishing various life transitions’ (p. 205). It also refers to the orderly
sequencing of life course transitions. Unlike in many contemporary western
societies, in Japan the hegemonic order and sequence of life transitions from
leaving the parental home, completing school (i.e. higher education), starting a
first job, and marrying accounts for over 90 percent of each cohort (p. 206).
Hareven adds that only after the Second World War did people in the USA
begin to adhere to more strict social rules regarding age norms, rather than post-
poning life transitions due to familial needs. As societies shift from agrarian and
pre-modern forms to modernity, Young (1988) has argued, the strength of the
natural solar, lunar and seasonal rhythms are being replaced by the symbolic
systems and timetables created by recent human culture. This separation of
the natural (identitaire, embedded) forms of the celestial realm and symbolic
(imaginary) systems and timetables results in the ‘metronomic society’ (Young,
1988). The transformation of temporality by cultural and social systems is often
realized through institutional calendars.
Institutional calendars
Institutional time, viewed through institutional calendars, is defined as the
timetables produced via the adaptation and life span of institutions. Institutions
are located within and produced by historical, social and cultural contexts.
Institutions in this sense are divided along religious, family, educational,
work/production, political, governmental and gender lines. These institutions
are often accompanied by seemingly innocent calendars that in fact often
regulate everyday life and thus most behaviour in the life course. Furthermore,
due to an individual’s position in these institutions, such as being male or
female, worker or non-worker, she or he experiences time in a different way.
These multiple calendars are often neglected, underestimated or viewed only in
07_Mills 2/2/2000 2:55 pm Page 104
relation to the family or production systems. The ontological effect of institu-

tional temporality is generally realized through shaping epistemological
approaches which examine the influence of point-in-time events such as
changes in divorce, abortion or parental leave legislation. For this reason, insti-
tutional calendars are now described in greater detail.
Calendars that are created and located within a specific historical-social-
cultural context dictate most individual action and group interaction in everyday
life (Fraser, 1987). Adam (1995) argues that this form of temporality mediates
most of our social interaction. The calendar, therefore, may also serve as a basis
for group solidarity and identity, or for the ‘mechanical solidarity’ of a society
(Zerubavel, 1982). Calendars can be agrarian, Julian, Mayan, Chinese lunar
zodiac, Hindu Mahabharata, Catholic liturgical, Arabic lunar, Coptic, family,
education, work, political, sports, or many others. The origin of the calendar is
often an interweaving of natural, cosmic and agricultural cycles with socially
constructed aspects of religious cosmologies.
Religious institutions are closely connected to the creation of calendars that
influence everyday life. These calendars enforce the observance of sacred times
such as Ramadan or the Sabbath, and rituals that mark phases in the life course
(e.g. baptism, bar mitzvah, wedding, funeral) (Zerubavel, 1981). They also
determine auspicious and inauspicious times or years for life events such as
marriage or childbirth. Two studies, for instance, have shown the historical
ecclesiastical restriction of marriage by the Catholic Church during Lent and
Advent (March and December) in The Netherlands and Belgium during the 17th
and 18th centuries (Van Poppel, 1995; Lesthaeghe, 1989). Similarily, Trovato
and Odynak (1993) attributed the peak of September births in Canada not only
to climate or natural causes, but also to the festival effect of Christmas and over-
all individual preference.
The Chinese lunar calendar is an interesting example of how this blend of
the natural lunar, cultural and religious cosmology influences partnership and
fertility behaviour. The calendar has 12 lunar months, each lasting 29 to 30
days, and one new moon (embolism) is added every two to three years to match
the western calendar (Abeysinghe, 1991).4 The timing of demographic events
has been linked to the Chinese New Year and auspicious and inauspicious years
within the 12-year lunar cycle. Goodkind (1996) connected the timing of births
among Chinese in Singapore between 1970 and 1990 to this calendar. He dis-
covered a ‘baby boom’ in the auspicious year of the Dragon and a ‘bust’ in the
inauspicious year of the Tiger. Thang and Swenson (1996) also found a correla-
tion in Vietnam between births and marriages in years considered auspicious in
the Chinese and Vietnamese astrological calendars. They concluded that indi-
viduals planned marriage and childbirth to coincide with optimal calendar
timing. Abeysinghe (1991) also found that marriages peaked during the Chinese
New Year in Malaysia, Singapore, Hong Kong and Taiwan. An additional
07_Mills 2/2/2000 2:55 pm Page 105
consideration is artificial peaks and drops attributed to the timing of official

registration of events. We will return to this important factor in section 2.
The influence of the western or Gregorian calendar, which marks the turn of
the millennium, is another excellent example of how something as arbitrary as
the date on a calendar can influence life course behaviour. The transition to the
new millennium is viewed as an important event in many cultures. It has also
created anxiety due to the risk of technical chaos in the form of the ‘millennium
bug’. However, it also serves as an impetus for retrospection about time itself,
the nature of society and, as any individual who has ever made a ‘New Year’s
Resolution’ knows, reflexive thinking about our own lifestyle and life course.
As Giddens (1994: 56) summates:
We live, as everyone knows, at a time of endings. There is, first of all, the end not
just of a century but of a millennium: something which has no content, and which
is wholly arbitrary – a date on a calendar – has such a power of reification that it
holds us in thrall. Fin de siècle has become widely identified with feelings of dis-
orientation and malaise, to such a degree that one might wonder whether all the
talk of endings, such as the end of modernity, or the end of history, simply reflects
them.
The turn of the century, indexed by a calendar, not only causes anxiety as
we realize that we rely upon all-pervasive technology, but also forces us to re-
evaluate our entire ontological construction of the world. The rise in numbers of
books about calendars and time, and the construction of monuments such as the
Millennium Dome in the UK, add to a feeling that a calendar event is indeed
momentous.
This calendar-derived event may also influence fertility and partnership
behaviour in the form of a mini-‘baby boom’ or rise in marriages. Although at
the time of writing it is too soon for researchers to examine any evidence, the
printed and electronic media are full of examples, including a ‘Millennium
Mom’ web page on the Internet which gives potential parents instructions about
how to time a pregnancy, with 9 April 1999 suggested as the optimum date for
conceiving the first baby to enter the new millennium. One potential
Millennium Mom planned the event to the hour, using a pregnancy planning kit
with ovulation predictors. As she explains in her own temporally laden words:
Most women ovulate between 24–36 hours after the LH [luteinizing hormone that
triggers ovulation] surge is detected. I used a stick this morning (2:00 a.m.) and
the test was positive! Calculate this for yourself and we are dead on for 4/9!5
This echoes the previous discussion regarding the colonization of our natural
biological cycles by new technologies. In response to fertility planning, and
added festivities, many hospitals, police and others were warned of a holiday
ban at the turn of the millennium in many countries (see source in note 5).
07_Mills 2/2/2000 2:55 pm Page 106
The new millennium may also cause a rise in marriages. Reuters news service
(5 May 1999) reported that the New Zealand parliament has relaxed a rule
banning weddings between 10 p.m. and 6 a.m. in order that marriages may
occur in the first time zone to enter the new century. The government-backed
‘First to the Future’ campaign capitalizes on the influence of this calendar date.
The important point is that a date or a year in a calendar that appears to be
insignificant at face value may have direct consequences on life event planning.
Furthermore, people’s interpretations or experiences of this calendar pervade
their everyday life activities. Just as religion has determined many calendars,
another institution central to the construction of events in the life course has
remained and evolved through time. This is the institution of the family.
Family time is another common topic in life course research (Hareven, 1994;
Ryder, 1992). It is recognized that individuals live and make decisions in the
context of families: first as children, then as partners and in-laws, and then as
parents. Family time includes aspects of the life cycle of families and house-
holds, spousal time (e.g. dual-earner couples) and parent–child times (e.g.
increasing participation of fathers). Ryder maintains that ‘the family is above all
the institution to which is assigned the responsibility for attempting to solve
problems of the passage of time both for the individual and for the population’
(1992: 161). In an applied study, Young (1992) focused on the interdependent
lives of partners and family members by considering the layers of individual,
family and historical time in Malaysia. For instance, the migration or death of
the head of the household results in a drastic change for the interdependent
family members. Hareven (1994) also clearly connects the levels of the indi-
vidual, family and historical time. Life history data can generally be used to
examine the individual, family or household. However, as the influence of
more traditional institutions such as the family and religion wanes in many
western societies, timetables are being increasingly shaped by other forms of
temporality beyond the sacred or traditional (Lesthaeghe, 1989). This ‘civiliza-
tion’ of time means that daily life becomes centred around more industrialized
cycles of education, work, politics, economics, sports or other secular time-
tables (Nishimoto Ikuko, 1997).
Educational calendars shape children’s identities, arguably training them for
entry into the next social institution of the labour force (Adam, 1995). It is the
first space in which individuals become aware of bureaucratic or commodified
time. This includes the importance of punctuality and being ‘on time’ in grade
progression, which will again occur in relation to partnerships and fertility. This
is also the first experience of distinct monotemporal divisions of time allotted
to specified subjects. When these children enter adulthood, they will find
(especially the women) that this monotemporal division of time cannot
accommodate the interdependent life domains of partnership, fertility, educa-
tion and work. In addition, the ‘proper’ timing of the transition from education
07_Mills 2/2/2000 2:55 pm Page 107
to work calendar has a pervasive influence in shaping further life course transi-
tions such as marriage or migration.
The work or production calendar regulates many everyday activities and
forces us to focus on time-management and setting deadlines. Individuals must
also timetable their leisure or private family time around their public employ-
ment time. The importance of agricultural and production cycles in relation to
the timing of marriages is discussed by Gunn (1990). There has been a general
shift to production time in what Van der Poel (1997) terms the ‘modularization’,
or commodification, of time, which is abstract and exchangeable. The rise of
temporal reflexivity (timetabling, planning) and construction of one’s own life
trajectory in western societies produces a surplus of a new type of time –
leisure time (Van der Poel, 1997). There is a division not only between private
and public time, but also between leisure or disposable time. As Giddens (1991:
77) states, ‘self-actualisation implies the control of time – essentially, the estab-
lishing of zones of personal time which have only remote connections with
external temporal orders’. An individual’s ability to focus on the development
of self-actualization stems, however, from her or his historical and geographic
location, family membership, sex and worker status. Relative integration into
the production calendar has been largely gender based, and thus the construction
of production, reproduction and gendered time has to be discussed together in
terms of social accessibility.
Gender-based calendars, or (wo)men’s temporal experience, have been
surprisingly absent from many demographic and life course studies. (Wo)men’s
time refers to the disparate temporal experience and use of time by women and
men. There is a history of divisions between the sexes in terms of reproduction
and production. The most obvious difference in time use by men and women is
reproductive or ‘procreation’ time (Tobler, 1996). However, there are also
labour differences, such as women doing more part-time work or being allotted
‘spaces’ in which to act (Friberg, 1993). These disparate temporal calendars are
rarely brought together or seen as defining factors, but instead remain largely
studies of women’s time (Kristeva, 1981; Davies, 1989; Forman, 1989; Friberg,
1993; Leccardi, 1996). When fertility is discussed, women are often the
subjects, yet the unique female experience of space and time is ironically
forgotten, or at worst, women are linked solely to reproduction (Tobler, 1996).
Women’s and men’s perceptions and uses of time often differ substantially,
ranging from basic biological differences to distinct patterns of managing
public and private time and space. These calendars are now discussed in relation
to work or production time.
An individual’s social accessibility is centred on the division between private
and public time. Private time is ‘deliberately designed to prevent, or at least
discourage, the formation of human contact and to separate people from one
another’ (Zerubavel, 1981: 143). Public time ‘is deliberately meant to promote
07_Mills
108
FIGURE 1
Unpaid and paid work hours of women and men in selected countries
2/2/2000 2:55 pm
Page 108
TIME
&
SOCIETY
9(1)
Source: United Nations (1996: Social Statistics Indicators, Series K, No. 12, p. 132). Japan (1991), Israel (1991-2), Former USSR (1986), Netherlands
(1988), Canada (1992).
07_Mills 2/2/2000 2:55 pm Page 109
the establishment of human contact and draw people together’ (Zerubavel,

1981: 143; see also Leccardi, 1996). In a study of women’s everyday lives,
Friberg (1993) observed that many of the structures and amenities that exist in
public space are male dominated, leaving private household space to women.
Figure 1 provides an illustration of the temporal division by sex, showing the
number of paid (public) and unpaid (private) hours currently worked by women
and men in different regions of the world. To varying degrees when public and
private work are accounted for, women engage in more ‘productive’ hours per
week than their male counterparts. The largest differences are apparent in the
division of unpaid private labour time. Women consistently work nearly double
the private time in Canada, The Netherlands and the former USSR, triple the
time in Israel and almost 10 times more in Japan (United Nations, 1996: 132).
This is largely attributed to the private–public gender space division, which
means that women spend more private time on household chores, preparing
meals, childcare and shopping. These long-term patterns of gender division and
segregation of public and private time in turn gain political significance and
reinforce themselves (Leccardi, 1996: 175). This time conflict is increasingly
addressed in studies of women’s parallel life course, such as the combination of
motherhood and employment.
The epistemological approach taken by researchers may entail making
ontological assumptions of temporality. The researcher may make natural or
cosmic cycles, biological or psychological development, historical period or
institutional calendars created by cultural and social time central to the analysis.
However, temporality may not be the sole influence on the supposed ‘temporal
effects’ that we observe in life histories. Up to this point, the discussion has
intentionally excluded any detailed mention of the data, methods and metho-
dology used to study the life course. In the previous section, it was argued that
temporality in relation to the natural or biological determined context should be
separated from what is symbolically, socially and culturally constructed. The
next section now builds on this division by addressing the interplay and
necessary separation between what is a temporal effect and what is related to
errors, anomalies and approaches in data, methods and methodology.
2. Temporal Aspects of Data, Methods and Methodology
Data, methods and methodology are the epistemological instruments linking

empirical evidence to the social processes that make up the life course. This
section argues that as researchers we should clearly isolate what is an effect of
temporality from what is an artefact of the epistemological approach of data,
methods and/or methodology. Life history data are generally taken from
surveys or statistics, parish or medical records. Surveys may be prospective or
07_Mills 2/2/2000 2:55 pm Page 110
retrospective. Prospective, or panel data, are collected at discrete points in time,

often following the same group of individuals. Retrospective cross-sectional
surveys gather information on the timing of recorded past life histories of indi-
viduals. Although the survey is taken at one point in time, the life history of an
individual is reconstructed by distinguishing the time and sequence of life
events. Methods are the techniques used to gather evidence and acquire knowl-
edge, such as qualitative methods of interviewing or observation, or quantitative
approaches discussed here.
Four distinctions between the data, method and temporality are discussed
initially. These are timing of data collection, memory or recall problems,
method of data gathering and self-registration. The first three are related to
survey data, which is most often used in event history analyses of the life
course. The last is related to vital statistics or other largely respondent-regulated
records of life events. The discussion then turns to innovations in methodology
that have been developed to deal with temporal problems. Methodology is the
way that the ontological or theoretical structure finds its application in the
research design. A common approach to study of the life course, for instance, is
the use of age, cohort and period within statistical techniques such as regression
models. It is concluded that although some temporal problems are already
recognized, several aspects are still lacking. These are reflexivity at all points in
research, new ways of understanding the memory of life events, blending of
qualitative and external data sources, and the use of methodology that expressly
deals with temporal assumptions.
Timing of data collection: selectivity and age truncation

The timing of data collection is important to ensure that the survey sample is
representative of the actual population. The first basic pitfall is that a survey
collects data only from survivors and non-migrants. For instance, in areas with
many deaths due to war or economic or political migration, the researcher must
realize who is not in the sample due to the influence of historical time. Second,
life history data are often only collected in the labour productive (18–65) or
fertility reproductive (15–49) periods of individual lives. The institutionally
defined ontological assumptions that place the (re)productive periods as the
most important times in the life course dictate these ages for data collection.
Arguably, many major life transitions related to education, partnership, fertility
and work take place within these age intervals. However, attention to cumula-
tive lifetime experience, such as childhood experiences (e.g. divorce, unem-
ployment, migration of parents) could also be determinants of decisions in the
adult life course. When postponement of life events and the advance of repro-
ductive and life-saving technologies are taken into account, classic age intervals
(e.g. 15–49) of life events and data collection will be increasingly eroded.
07_Mills 2/2/2000 2:55 pm Page 111
Furthermore, as the ‘baby boom’ cohort in many western regions begins to age,
this cohort effect will produce a new interest in later life course events such as
grandparenthood, health and active life expectancy. Finally, the assumption that
women are central to reproduction results in the epistemological manifestation
of fertility data gathered only from women, thus omitting the unique fertility
histories of men.
A third aspect related to the timing of data collection is age truncation, which
addresses the problem of when the data are collected in the progression of
the respondent’s life course. This is a common problem when examining life
histories of younger individuals. Since the survey is taken at a particular time in
their life course, younger people have not had enough time to accumulate
higher order life transitions by the date of the survey (e.g. second child, widow-
hood). In this case, the age of the individual at the date of the survey is a distinct
methodological factor in defining the partnership or fertility process. A Lexis
diagram is useful in understanding the differential amount of time lived or
observed by different cohort or age groups.6
Figure 2 demonstrates that cross-sectional retrospective survey data provide
only a partial information set of life history data. The diagram shows a sample
of individuals born from 1943 to 1980, aged 13 to 49 at the date of survey and
divided into four cohort groups. It is easy to see from this figure that we have
life history information on a disparate number of years for each cohort of
women. For the oldest cohort, born from 1943 to 1949, we have life history
information up to age 42 to 49. However, for the youngest cohort, born from
1970 to 1980, we have life history information until approximately age 13 to 22.
We do not obtain a detailed life history of the younger respondents, as many of
their life events will be censored (not occurred) by the date of the survey. We
may not have detailed information for middle cohorts either. For instance, many
younger women in western European countries are postponing life events such
as entry into a partnership or motherhood in comparison to older cohorts or
generations. In addition, since the survey is only collected up to age 49, we are
unlikely to obtain information about higher order union experiences, such as
partnership dissolution due to death of the partner. Another temporally related
aspect apparent in Figure 2 is the multiple clocks that are used to study the indi-
vidual’s life course. These are discussed in detail in the methodological section
(p. 118).
In addition to the importance of the timing of data collection in the indi-
vidual’s personal clock, the timing in the day, month or year that data are
gathered can have implications for results. The number or monthly variation in
specified days, such as variations by dates of religious festivals or national
holidays, can influence the data. ‘Seasonally adjusted rates’ are often used, for
instance, to reflect seasonal employment opportunities and activity levels.
07_Mills 2/2/2000 2:55 pm Page 112
FIGURE 2
Lexis diagram of exposure time by birth cohort, year of birth and age at the time of
survey in years and Century Month Code units (1993)
Memory problems
Another common problem in retrospective data collection is that of memory or
recall of past life events, which results in misreporting. Memory problems
are often connected with age, sex, social and cultural conceptions of time. When
gathering life history data, researchers must not only be aware of the
respondent’s experience of time, but also reflexively examine their own
temporal presumptions. Data problems such as age heaping or digit preference
related to cultural mismatches of respondent–researcher visions of time are
discussed first. However, missing or misreported temporal data can also be
compared by the type of life event. The second point suggests that the meaning
or importance that one attaches to family or life course events plays an
additional role in remembering the date of the event. Furthermore, it is not only
the individual’s conception, but also the ritualization of the event in the current
historical-cultural-social context that affects memory.
A common problem associated with recall and misdating often manifests
itself in the form of age heaping or digit preference, resulting in anomalies such
07_Mills 2/2/2000 2:55 pm Page 113
FIGURE 3
Age heaping in reported age at first marriage, women, Uttar Pradesh, India
Source: National Family and Health Survey, 1992–3.
as over-representation of dates or ages ending in zero or five. This is attributed

to the fact that in many contexts, age or the date of an event by month or year is
not an intuitive indicator of time. Figure 3 is a clear example of age heaping and
the need to be wary of ontological perceptions of temporality. This figure shows
the reported age at first marriage for women in the Indian state of Uttar Pradesh
using the 1992–3 National Family and Health Survey. Digit preference has
clearly occurred at age 5, 10, 15 and 20 years. However, socially (or perhaps
legally) elevated heaping may also occur at age 18, which is the legal age of
marriage for women in India. The data may reflect either more marriages at age
18, or the socially constructed data artefact of ‘reported’ marriages at this age.
Numerous applications are based solely on the adjusting, imputing, indexing
and standardizing of age, and various mathematical indices have been
developed to ‘smooth’ or account for the extent of misrepresentations such as
these.7 A better solution, however, would be to start at the root of the problem,
which is in the first phase of research questions, questionnaire construction and
data gathering. This is discussed in the next section.
In addition to age heaping, certain types of temporal data may be missing or
misreported, and such data also vary by the type of respondent and life event.
When examining the extent of misreported or missing data, older respondents
07_Mills 2/2/2000 2:55 pm Page 114
typically provide less accurate responses. However, aspects beyond the age of
the respondent also affect memory. For instance: do women remember family
events better than men? Are there certain life events that have more missing
temporal data than others? In life history analysis, the month and year of a life
event are often the central indicators that specify the beginning and end of a life
stage, timing of transitions and mean age at occurrence of events. However, it is
common that for some types of life events, more respondents report the year,
but not the month of the event. Poulain et al. (1991) compared the accuracy of
temporal reporting of life events by couples. They confirmed differences in
reporting by sex and type of event. A sample of 445 couples was first inter-
viewed in separate rooms about family-related life history events. They were
then allowed to adjust for any uncertain dates via the use of family records, rent
receipts, etc. The researchers then consulted life statistics registers to check
the accuracy of temporal reporting. They found that dates were satisfactorily
reported for marriage and birth or death of a child, dates also traceable by regis-
tration records. Marriage and birth of children were reported within a month in
90 percent of the cases. However, migration of the household or date of children
leaving the parental home were only exact to within a month for 39–67 percent
of the cases. Finally, the information provided by women was more reliable
than by men, with the combined spousal answers being more accurate again.
In the study of partnership formation, data for the month of cohabiting unions
are more likely to be missing than for legal marriages. In the 1993 Netherlands
Fertility and Family Survey, for instance, there is a difference in missing data
for the month of event by type of union and sex. Table 1 shows the missing
month dates for unions by frequency and percentage for women and men.
Percentages with a small number of cases should be regarded with caution.
Despite this, it is still possible to see a general difference between the
higher percentage of persons who cannot recall the beginning month of their
cohabiting or consensual partnerships, particularly for males. When comparing
first union formation, around 0.3 to 0.4 percent of women and men do not recall
the month of their first marriage, compared to 7.8 percent who do not recall the
month of their first cohabitation. The amount of missing data is higher in later
unions. This could be attributed to the fact that less ritual or meaning is placed
on these repeated unions. Finally, comparing the percentages of women and
men who cannot recall the month of the partnership event, men generally have
more problems remembering the month, particularly union dissolution. This
basic comparison supports Poulain et al.’s (1991) findings (but should be
examined with more sophisticated methods). Beyond identifying the problem of
memory in data gathering, researchers have attempted to find mnemonic
devices and collection strategies to improve the quality of life history data.
07_Mills 2/2/2000 2:55 pm Page 115
TABLE 1
Frequency and percentage of reported and missing cases in the monthly timing of
union events by sex and type of union
Women (n=4516) Men (n=3705)
Month of event Month of event
Union Event Reported Missing (%) Reported Missing (%)
Marital unions
First marriage 2632 7 (0.3) 1745 7 (0.4)
First dissolution 314 25 (8.0) 177 15 (8.5)
Second marriage 114 2 (1.8) 68 2 (2.9)
Second dissolution 13 1 (7.7) 8 4 (50.0)
Cohabiting unions
First union 961 75 (7.8) 741 58 (7.8)
First dissolution 441 48 (10.9) 352 34 (9.7)
Second union 149 5 (3.4) 144 11 (7.6)
Second dissolution 153 3 (1.9) 57 10 (17.5)
Third union 16 – – 26 – –
Third dissolution 7 – – 8 1 (12.5)
Source: 1993 Netherlands Fertility and Family Survey [Onderzoek Gezinsvorming].
Note: Totals may differ slightly due to weighting. Percentages are calculated from the valid number
of year of reported events.
Data gathering: understanding the ontological experience of temporality

In recent years, solutions have targeted the root of memory problems, rather
than focusing on post-survey adjustment of the data. In 1988, Freedman et al.
developed a life history calendar now widely used by researchers in the field.
They argue that using this calendar improves the quality of retrospective data
collection because it facilitates visual and mnemonic timing of various life
events. In the first step of constructing this life history, a calendar-based
framework is utilized to provide a basic reference point for events. The more
significant and thus easily remembered events such as marriages, births and
deaths, are asked for initially to serve as temporal or mnemonic markers for
other questions. After this skeleton of life course events is established, more
detailed occurrence, timing and sequencing of events are more readily recalled.
Hertrich (1993) also provides an explicit discussion and solution to age estima-
tion problems in her discussion of data collection in a rural African community
in Mali. Beyond chronological classification of events and more technical esti-
mation methods, she argued that existing sources such as registrars’ records,
Christian missions and maternity records could be used to resolve dating
07_Mills 2/2/2000 2:55 pm Page 116
problems in survey information. These events can be used as ‘milestones’ or

‘dating references’ for individual or family history construction.
The discussion to this point has been based on problems related mostly to
survey data where the researcher sets the agenda in the data gathering process.
The degree to which researchers examine their own conceptions of time in addi-
tion to that of the respondent is essential in separating what is truly temporal and
what is related to data error. However, life histories can also be constructed
from historical, parish, medical or registrar’s records where events are self-
registered or controlled by the registration agency. This leads us to the final
problem, which is the influence of data registration on temporal data.
Self-registration: the influence of data collection on temporality

In some cases, phenomena that have been associated with temporal effects such
as seasonality or the influence of institutional calendars may in fact be attributed
to the data. Two examples of the monthly variation of timing of events are
shown in Figure 4. The two lines represent the average monthly percentage
distribution of first marriages in China from 1950 to 1981 and births in the
former USSR from 1958 to 1985. A striking difference is that 20 percent of first
marriages in China occur in January and 14 percent in December. This supports
the previous discussion that marriages peak during the Chinese New Year in
January and February. This ‘festival effect’ should not, however, result in
a higher number of marriages in October (9.9 %), November (9.8 %) and
December (14 %) as Figure 4 shows. We can conclude from this, with
Abeysinghe (1991), that the jump in marriages in the latter part of the year can
be attributed to early registration of marriages spurred on by the festival effect.
If first marriages were explained without acknowledging the force of both regis-
tration and institutional calendars, results would be highly misleading.
Figure 4 also shows monthly distribution of the mean daily number of births
reported by calendar month in the former USSR. Anderson and Silver (1988:
306) found that there was a substantial drop in births in December (of 12 %) and
an excess of births in January (of 9 %) that could not be attributed to any institu-
tional, natural or biological calendar. After a detailed investigation and analysis
of the data, they attributed these seasonal fluctuations to changes in the
country’s registration system more than to any true seasonal effects. They found
evidence that many births that actually occurred in December were officially
reported as occurring in January of the next year. This they attributed to the fact
that Soviet law permits a 30-day delay between a birth and its registration.
People tended to delay the registration of life events during the December
holiday period. Finally, they found that registration offices conducted their
yearly audit to uncover missing data and unreported births in January of each
year. These missing births were then added to the January records. These find-
07_Mills 2/2/2000 2:55 pm Page 117
FIGURE 4
Average monthly percentage of distribution of events:
first marriages in China, 1950-81 and births in USSR, 1958-85
Source: Xuan (1983: 107) in Yi (1985: 33), ‘Average monthly percentage of distribution of first
marriage in China, 1950–81’, and calculated from Anderson and Silver (1988: 307) ‘Mean
daily number of births reported by calendar month, for USSR, 1958–85’.
ings show that the institutional calendar via the festival effect, for instance, may
affect not only the occurrence, but the registration of demographic events.
Appreciating the full effects of methods of data collection and understanding
the way that respondents and researchers perceive time are increasingly impor-
tant features of quantitative life course research. However, data gathering is
only the first step in this epistemological approach. The second step is the
methodological undertaking of data analysis, frequently using life table and
regression models.
Methodology: the force of ontology on the epistemological approach

Methodology in this field is often based on the ontological assumptions already
outlined: natural, human development, historical, and social time. The analysis
therefore integrates the clocks of age, period, cohort, duration, or episode
(parity, partnership, job). When natural or cosmic temporality is considered as
the underlying factor governing the occurrence of events, the epistemological
approach often centres on seasonality. If biological or human development is
07_Mills 2/2/2000 2:55 pm Page 118
seen as a primary factor shaping the life course, the social gauge of age progres-
sion is often used to judge if someone is ‘on’ or ‘off’ time. The ontological
effects of shared temporal mentalities are often studied by placing cohort or
period as central in the analysis. Measures of social change are frequently deter-
mined by comparing birth cohorts, a method which ties macro-level events (e.g.
economic depression, postwar period) to micro-level individual and meso-level
family adaptations and outcomes (Ryder, 1965; Elder, 1974, 1991, 1994;
Hogan, 1981). Conceptually and methodologically, it is a challenge to dis-
entangle the effects of age, cohort and period. Age effects are related to growing
older, such as accruing higher education or income over time. Cohort effects are
common to people born at the same time, such as historical events, oppor-
tunities, constraints, competition and size. Period effects are the result of this
shared experience of historical events, or historical time. The choice of which of
these time axes to study rests firmly on the researcher’s shoulders and depends
upon theoretical and statistical needs (Blossfeld and Rohwer, 1995).
A variety of clocks are often combined to reconstruct the life course after the
timing of life events has been gathered or registered. Recall that Figure 2 shows
a blend of five different methodological clocks. The left and bottom axes show
a time unit commonly used in analysis, Century Month Codes (CMCs). CMCs
are the number of months since the beginning of the 20th century. For instance,
January 1900 has a CMC value of one, since it is one month since the beginning
of the century. A CMC of 555 would be March 1946, and so on. These units are
often used in event history analysis of the life course, which is a broad term for
longitudinal statistical techniques. This approach investigates transitions,
timing, duration, and sequencing of events in the life course (Tuma and Hannan,
1984; Mayer and Tuma, 1990; Courgeau and Lelièvre, 1992; Blossfeld and
Rohwer, 1995). Using CMCs, chronological time can be standardized on an
identical clock for each respondent, regardless of age, life stage, and historical
period or cohort membership. The first CMC clock represents the age of the
respondent in relation to the turn of the century. This corresponds to the second
clock, which is the respondent’s biological or individual age in years on the
right axis of the diagram. The third clock, represented by the diagonal lines,
shows the four cohort divisions that can be used to compare social change or
differences in life course behaviour over historical time. The fourth and fifth
clocks, shown at the bottom of the diagram, are historical (period) times in years
and CMC units since the turn of the century.
Another common way to organize life history data is by episodes (e.g. first,
second, third birth episode) within each life career. This relates to another
measure often referred to as process time (Blossfeld, 1996; Courgeau and
Lelièvre, 1992; De Bruijn, 1999; Willekens, 1997). Process time is the length of
time since the last demographic event of interest (e.g. first birth) and the onset of
the process when the individual becomes ‘at risk’ of experiencing another event
07_Mills 2/2/2000 2:55 pm Page 119
(e.g. second birth). This type of time is frequently related to survival, duration
and Semi-Markov modeling and analysis. Adam (1995: 21) likewise stresses
interplay between clock and calendar time and the timing of events and social
processes: ‘the existence of clock and calendar time does not prevent us from
locating the past, present and future with reference to events, processes and
social relations’. Based on what is seen as theoretically or substantively impor-
tant in the analysis, different or combined clocks such as age (in years or
CMCs), cohort, historical period, generation, or process time are used as tempo-
ral indicators in analysis. However, information on life events is often reported
in the present, as the individual recalls the past and imagines the future. As
discussed in the opening section of this article, the life-world of the present is
often interpenetrated by these temporal realms.
The past, the present and the future

The notion that the present contains remnants of the past and pages of the future
is increasingly mentioned by researchers in this field (De Gans, 1994). Life
course outcomes are seen as embedded in a process that follows a unique narra-
tive path (Willekens, 1990). Proponents of event history analysis, such as
Courgeau and Lelièvre (1992: 61) maintain that ‘the hypothesis which consists
in ignoring the influence of the past on the future development of a process
leads to errors of appreciation or interpretation of the phenomena observed’. De
Gans (1994: 339) argues that after an event occurs there are ‘adjustment
responses’ which ‘affect the understanding and the image of the past and have a
modifying effect on goals and plans for the future’. Mayer and Tuma (1990: 7)
likewise argue that ‘life events and phases must be studied as part of a life
trajectory in which later outcomes are partly consequences of earlier conditions,
events, and experiences’. If this is indeed the case, how do we incorporate these
ontological assumptions into quantitative research?
One assumption in life history analysis is that an underlying stochastic
process generates life course events (Willekens, 1987). Stochastic processes are
a collection of random variables that chart the evolution of interdependent
processes over time (Tuma and Hannan, 1984). A random variable is defined
by a set of possible values it may take and an associated relative frequency
(probability) function (Namboodiri, 1991: 17). In simple notation, S refers to the
discrete state space and X(t) is the random variable, or the state occupied at time
t (>0). Each variable contains a systematic part related to the process and a
random part related to the distribution, introducing chance and probabilistic
predictions. Many models are Markov-based and assume that event occurrence
depends on the origin state (e.g. divorced) and overlook the process history.
How an individual reaches a divorced state, for instance, is often contingent
on the sequencing, type and duration of previous life events (e.g. first cohabited,
07_Mills 2/2/2000 2:55 pm Page 120
duration of marriage) and interaction with other life course careers (e.g.
presence of children, gainful employment). If duration is considered important,
the model can be extended to a semi-Markov. This reflects not only the occur-
rence of the life transition, but also the importance of duration, seniority or
experience in a state, often referred to as duration-dependence (Namboodiri,
1991; Willekens, 1997). This duration between two states defines a stage in an
underlying developmental process (Chiang, 1984: 246 in Willekens, 1997). A
further extension of these models is to consider the entire life history or
sequencing of life events of the individual in a non-Markov model (Rajulton,
1992). This type of model would be the epistemological equivalent to the onto-
logical assumptions of the importance of the past life path (path-dependence) on
the present and future life course of the individual. It can include information on
the timing, but more important, the sequencing and number of events. In this
way, primary sequential patterns of life course paths are determined.
A dynamic aspect of time can also be introduced into the regression model
(for a detailed discussion, see Blossfeld, 1996: 193–6). Explanatory factors or
covariates are used to describe the impact of certain characteristics (education,
residence) on the life event probability or rate. Covariates may be fixed, i.e.
remain the same over the entire life span (sex, place of birth) or time-varying,
i.e. changing values over time (education level, socio-economic status). When
examining the first birth experiences of women born in 1946, for instance, the
covariate influence does not lie in their current education, residence or socio-
economic status at the time of the survey. Rather, it depends on the values of
these variables at the time when they were ‘at risk’ of experiencing the demo-
graphic event (e.g. education status at time of marriage).
Finally, understanding how individuals plan their future life course in light of
the present and probabilistic future takes us into the field of projections and
micro-simulation. In this type of analysis, life history transition probabilities
from the present are used to simulate or project future partnership, fertility, and
workforce or migration scenarios of the future. Based on alternative scenarios
and on current developments in rates of the social processes under study (e.g.
marriage, divorce, fertility rates), a model is projected. This model then poten-
tially estimates the future life courses of individuals, households or population
groups (Willekens, 1990).
Conclusion
Although there have been innovations to incorporate temporality in life course

research, this discussion reveals that several areas are still wanting. The first is
the lack of reflexivity in research at all points of the analysis. The experience of
time by both the researcher and the respondent should be central when collect-
07_Mills 2/2/2000 2:55 pm Page 121
ing temporal data. Second, memory problems should be examined not only for
the timing of events, but for the extent of missing temporal data for types of
events and specific respondents. Third, quantitative life course research should
not only be ‘quantitative’. To fully realize ways that time is perceived and
reported, prior qualitative or ethnographic research may be necessary. This
means focusing on respondents’ and researchers’ everyday perceptions and
assumptions regarding time and the life course. From this ontological under-
standing, a more efficient and reliable epistemological tool, such as an ethno-
graphic approach or life history calendar, can be honed for each context. In
addition to qualitative research, external sources of data, such as family or
registrar’s records, could be used to trigger respondents’ memories of the true
timing of events. Finally, although innovative methods of data collection and
analysis have been introduced, researchers too often rely on classic measure-
ment of age, cohort, period or clock and calendar time. The approach should
therefore move beyond ‘fixing’, or arbitrarily dividing the data into convenient
categories, to trying to understand the underlying interplay between tempo-
rality, our epistemological approach and the life course.
The main intent of this article was to heighten awareness of the ontological
impact of temporality for life course researchers. I have also tried to identify the
strengths and gaps, translating awareness into new research possibilities.
Temporality needs to be introduced at multiple levels from the cosmic to the
individual’s conscious cognition. The force of temporality is also present at all
points in research, from the reflexive perception of time in different contexts to
research hypotheses, data gathering and development of a statistical model. In
addition to presenting traditional temporal influences such as natural, biological
and psychological, historical and social time, aspects less often addressed by
quantitative approaches were presented. These included the importance of
cultural conceptions of time and distinct temporal experience by groups of indi-
viduals such as women and men or workers and non-workers. Furthermore, it is
essential to recognize explicitly that institutional calendars shape many of our
activities. Through a detailed exploration of these timetables, it is clear that
seemingly arbitrary dates, such as the year of the Tiger or the coming of the
millennium, might have a tangible influence on life course construction. The
religious, family, educational, production and gendered experience of institu-
tional time can likewise influence one’s social accessibility, which is centred on
the daily divisions of public and private time.
Equally, too rarely is the link made between theoretical notions of time and
the epistemological approach of the data and methodology. The timing of data
collection and the ontological reality of respondents and researchers require
unique data collection methods, such as ethnographic interviews, life history
calendars and use of external records to verify data or serve as mnemonic
devices. In addition, not only the sex or age of the respondent, but also the type
07_Mills 2/2/2000 2:55 pm Page 122
of demographic event may influence memory and subsequent reporting. There

is little research in how we remember different types of life events. Finally,
institutional calendars may also influence the registration of life events, such as
the pre-registration of marriages or postponement of registration of births over
the holiday season. Research can integrate the collection of data using various
sources and methods in addition to utilizing the full breadth of the statistical
techniques offered through recent applications in event history and stochastic
analysis.
Beyond a review of multiple temporal effects, several general trends relevant
to life course research were noted. First, previously natural, biological and
religious calendars that regulated everyday life are increasingly colonized by
social constructions of time. This colonization or civilization of temporality
starts at the most fixed element of natural cosmic cycles through the intro-
duction of technological advances such as air conditioners or heating. Techno-
logy also allows the social construction of biological menstrual cycles through
contraceptives, for example, in addition to the ability to control or time fertility.
New types of time, such as leisure time, may also influence lifestyles and thus
life planning. Second, the type of institutional calendar that shapes our everyday
life has also shifted from religious, family or agricultural calendars to more
modernized aspects of educational and industrialized production timetables.
This shift in temporal experience results in more individualistic thinking about
the timing of life events. Individuals increasingly make decisions based on the
preferences about when to have a child, rather than being constrained by the
natural cycles of climate or biology, or by traditional religious cosmologies, or
gendered and family time. Finally, although this article focuses only on the tem-
poral, many of the aspects discussed, such as cultural and social perceptions of
time, are developed within and have distinct or multilevel spatial locations. The
hope is that temporal awareness will lead to increased multi-temporal applica-
tions where time in the life course can be seen as more than a synonym of age,
cohort, period, or as something to be smoothed, adjusted or taken for granted.
Our assumptions about the influence of time will then shape a distinct new
research approach.
Notes
The research was made possible through a Doctoral Fellowship from the Social Sciences
and Humanities Research Council of Canada. I would also like to thank Frans Willekens,
the anonymous reviewers, and particularly Mike Crang for their helpful comments on an
earlier draft of this paper.
1. This discussion is in no way intended to negate the significance of qualitative bio-

graphical or narrative life history studies.
2. The study of partnerships includes legal marriages and consensual unions. Marriages
07_Mills 2/2/2000 2:55 pm Page 123
are defined by the reported or registered event of a wedding. Consensual unions are
unmarried, intimate sexual partners sharing the same household (i.e. distinct from
merely co-residence). Aspects of fertility relate to reported or registered births and
social, cultural and biological matters associated with sexuality and conception.
3. National Family Health Survey, 1992–93, initiated by the Ministry of Health and
Family Welfare in India and coordinated by the International Institute for Population
Sciences in Mumbai, India.
4. Nishimoto Ikuko (1997) described how in 1873 the Japanese adopted, as opposed to
accommodating, western time measurement by assuming the 24-hour day and
Gregorian calendar.
5. Taken from the Everything 2000 Millennium Mom website,
http://www.everything2000.com/life/millenniummoms
6. For a more detailed description of the Lexis diagram in relation to marital unions, see
Willekens (1987).
7. These include rectangularity, linearity, Whipple’s, Myers’ Blended (1940), Bachi
(1951), Carrier (1959), Ramachandran (1965), and the United Nations (1967) age-
accuracy Indices (in Shryock and Seigal, 1976: 116–18).
References
Abeysinghe, T. (1991) ‘A Seasonal Analysis of Chinese Births’, Journal of Applied

Statistics 18(2): 275–86.
Adam, B. (1990) Time and Social Theory. Cambridge: Polity Press.
Adam, B. (1995) Timewatch: The Social Analysis of Time. Cambridge: Polity Press.
Anderson, B.A. and Silver, B.D. (1988) ‘The Effects of the Registration System on the
Seasonality of Births: The Case of the Soviet Union’, Population Studies 42: 303–20.
Bantje, H.F. (1988) ‘Female Stress and Birth Seasonality in Tanzania’, Journal of
Biosocial Science 20(2): 195–202.
Blossfeld, H.P. (1996) ‘Macro-sociology, Rational Choice Theory, and Time: A
Theoretical Perspective on the Empirical Analysis of Social Processes’, European
Sociological Review 12(2): 181–206.
Blossfeld, H.P. and Rohwer, G. (1995) Techniques of Event History Modeling: New
Approaches to Causal Analysis. Mahwah: Lawrence Erlbaum.
Bongaarts, J. and Potter, R.G. (1983) Fertility, Biology and Behaviour. New York:
Academic Press.
Bronowski, J. (1974) The Ascent of Man. Boston, MA: Little, Brown.
Courgeau, D. and Lelièvre, E. (1992) Event History Analysis in Demography. Oxford:
Clarendon Press.
D’Andrade, R. (1995) The Development of Cognitive Anthropology. Cambridge:
Cambridge University Press.
Davies, K. (1989) Women and Time. Weaving the Strands of Everyday Life. Lund:
Grahns Boktryckeri.
De Bruijn, B. (1999) Foundations of Demographic Theory. PDOD Publication Series A,
Doctoral Dissertation, University of Groningen, The Netherlands.
De Gans, H. (1994) ‘Breaking into Time? Problems of Time in Demographic Fore-
casting’, in E. Vilquin (ed.) Le Temps et la démographie. Chaire Quetelet 1993, pp.
329–58. Louvain-la-Neuve: Academia/L’Harmattan.
07_Mills 2/2/2000 2:55 pm Page 124
Domingues, J.M. (1995) ‘Sociological Theory and the Space–Time Dimension of Social
Systems’, Time & Society 4(2): 233–50.
Elchardus, M. (1984) ‘Life Cycle and Life Course: The Scheduling and Temporal
Integration of Life’, in S. Feld and R. Lesthaeghe (eds) Population and Societal
Outlook, pp. 251–67. Brussels: Koning Boudewijnstichting.
Elder, G.H., Jr (1974) Children of the Great Depression. Chicago: The University of
Chicago Press.
Elder, G.H., Jr (1991) ‘Lives and Social Change’, in W.R. Heinz (ed.) Theoretical
Advances in Life Course Research, pp. 58–86. Weinheim: Deutscher Studien Verlag.
Elder, G.H., Jr (1994) ‘Time, Human Agency and Social Change: Perspectives on the
Life Course’, Social Psychology Quarterly 57(1): 4–15.
Elias, N. (1987/1992) Time: An Essay. Oxford: Blackwell.
Erikson, E.H. (1950/1963) Childhood and Society. London: Hogarth Press.
Foot, D.K. and Stoffman, D. (1996) Boom, Bust and Echo. Toronto: Macfarlane Walter
& Ross.
Forman, F.J., ed. (1989) Taking our Time: Feminist Perspectives on Temporality.
Oxford: Pergamon Press.
Foster, J. (1996) ‘Menstrual Time: The Sociocognitive Mapping of “The Menstrual
Cycle” ’, Sociological Forum 11(3): 523–47.
Fraser, J.T. (1987) Time – The Familiar Stranger. Amherst: University of Massachusetts
Press.
Freedman, D. et al. (1988) ‘The Life History Calendar: A Technique for Collecting
Retrospective Data’, Sociological Methodology 17: 37–68.
Friberg, T. (1993) Everyday Life. Women’s Adaptive Strategies in Time and Space.
Lund: Lund University Press.
Giddens, A. (1981) ‘Agency, Institution and Time-Space Analysis’, in K. Knorr-Cetina
and A.V. Cicourel (eds) Advances in Social Theory and Methodology: Toward an
Integration of Micro- and Macro-Sociologies, pp. 161–74. London: Routledge &
Kegan Paul.
Giddens, A. (1984) The Constitution of Society: Outline of the Theory of Structuration.
Cambridge: Polity Press.
Giddens, A. (1991) Modernity and Self-Identity: Self and Society in the Late Modern
Age. Cambridge: Polity Press.
Giddens, A. (1994) ‘Living in a Post-Traditional Society’, in U. Beck, A. Giddens and S.
Lash (eds) Reflexive Modernization, pp. 56–109. Cambridge: Polity Press.
Giele, J.Z. and Elder, G.H., Jr, eds (1998) Methods of Life Course Research. London:
Sage.
Goodkind, D. (1996) ‘Chinese Lunar Birth Timing in Singapore: New Concerns for
Child Quality amidst Multicultural Modernity’, Journal of Marriage and the Family
58(3): 784–95.
Gunn, P.A. (1990) ‘Productive Cycles and the Season of Marriage: A Critical Test’,
Journal of Interdisciplinary History 21(2): 217–43.
Hägerstrand, T. (1975) ‘Space, Time and Human Conditions’, in A. Karlquist et al. (eds)
Dynamic Allocations of Urban Space, pp. 3–14. Wermeads: Saxon House.
Hägerstrand, T. (1988) ‘Time and Culture’, in G. Kirsch, P. Nijkamp, and K.
Zimmerman (eds) The Formulation of Time Preferences in a Multidisciplinary
Perspective, pp. 33–42. Avebury: Gower Press.
Hareven, T.K. (1994) ‘Synchronizing Individual Time, Family Time and Historical
07_Mills 2/2/2000 2:55 pm Page 125
Time,’ in E. Vilquin (ed.) Le Temps et la Démographie. Chaire Quetelet 1993, pp.

203–18. Louvain-la-Neuve,: Academia/L’Harmattan.
Hertrich, V. (1993) ‘The Contribution of Existing Sources to the Dating of Events. A
Survey in Mali’s Bwa Country’, Population 5: 73–99.
Hogan, D.P. (1981) Transitions and Social Change: The Early Lives of American Men.
New York: Academic Press.
International Institute for Population Sciences (IIPS) (1996) India. Women and Children
in Focus. New Delhi: IIPS.
James, W.H. (1990) ‘Seasonal Variation in Human Births’, Journal of Biosocial Science
22(1): 113–19.
Kristeva, J. (1981) ‘Women’s Time’, Signs 7(1): 13–35.
Lam, D.A. and Miron, J.A. (1996) ‘Modeling Seasonality in Fecundability, Conceptions,
and Births’, Demography 31(2): 321–46.
Latten, J. and De Graaf, A. (1997) Fertility and Family Surveys in Countries of the ECE
Region. Standard Country Report. The Netherlands. United Nations Economic
Commission for Europe, Economic Studies No. 10c. New York/Geneva: United
Nations.
Leccardi, C. (1996) ‘Rethinking Social Time: Feminist Perspectives’, Time & Society
5(2): 169–86.
Lesthaeghe, R. (1989) Marriage Seasonality, Moral Control and Reproduction in
Belgium, 1600–1900. Interuniversity Programme in Demography, Working Paper No.
4. Brussels: Vrije Universiteit Brussel.
Luhmann, N. (1977/1982) ‘The Future Cannot Begin: Temporal Structures in Modern
Society’, in N. Luhmann, The Differentiation of Society, pp. 271–88. New York:
Columbia University Press.
Mayer, K.U. and Tuma, N.B. (1990) ‘Life Course Research and Event History Analysis:
An Overview’, in K.U. Mayer and N.B. Tuma (eds) Event History Analysis in Life
Course Research, pp. 3–22. Madison: University of Wisconsin Press.
Mead, G.H. (1932/1980) The Philosophy of the Present. Chicago: The University of
Chicago Press.
Mukherjee, S., Bhattacharya, B.N. and Singh, K.K. (1996) ‘Distribution of Time of First
Birth in Presence of Social Customs Regulating Physical Separation and Coital
Frequency’, Mathematical Biosciences 131(1): 1–21.
Namboodiri, K. (1991) Demographic Analysis. A Stochastic Approach. San Diego, CA:
Academic Press.
Neugarten, B. L. (1979) ‘Time, Age and the Life Cycle’, American Journal of Psychiatry
136(6): 887–94.
Neugarten, B. L. and Hagestadt, G. O. (1976) ‘Age and the Life Course’, in R. H.
Binstock and E. Shanas (eds) Handbook of Aging and the Social Sciences, pp. 33–56.
New York: Van Nostrand.
Nishimoto, Ikuko (1997) ‘The “Civilization” of Time: Japan and the Adoption of the
Western Time System’, Time & Society 6(2/3): 237–59.
Nowotny, H. (1992) ‘Time and Social Theory: Towards a Social Theory of Time’, Time
& Society 1(3): 421–54.
Nowotny, H. (1994) Time: The Modern and Postmodern Experience. Cambridge: Polity
Press.
Panter-Brick, C. (1996) ‘Proximate Determinants of Birth Seasonality and Conception
Failure in Nepal’, Population Studies 50(2): 203–20.
07_Mills 2/2/2000 2:55 pm Page 126
Poppel, F. van (1995) ‘Seasonality of Work, Religion and Popular Customs: The
Seasonality of Marriage in Nineteenth and Twentieth-century Netherlands’, Con-
tinuity and Change 10(2): 215–56.
Poulain, M., Riandey, B. and Firdion, J.M. (1991) ‘Enquete biographique et registre
belge de population: une confrontation de données’ [Life history surveys and popula-
tion registers in Belgium: A comparison of data], Population 46(1): 65–88.
Prigogine, I. (1980) From Being to Becoming. Time and Complexity in the Physical
Sciences. San Francisco, CA: Freeman & Co.
Rajulton, F. (1992) ‘Life History Analysis: Guidelines for Using the Program LIFE-
HIST’, Discussion Paper 92–5, Population Studies Centre, University of Western
Ontario, Canada.
Russell, D., Douglas, A.S. and Allan, T.M. (1993) ‘Changing Seasonality of Birth – A
Possible Environmental Effect’, Journal of Epidemiology and Community Health
47(5): 362–7.
Ryder, N.B. (1965) ‘The Cohort as a Concept in the Study of Social Change’, American
Sociological Review 20: 843–61.
Ryder, N.B. (1992) ‘The Centrality of Time in the Study of the Family’, in E. Berqúo and
P. Xenos (eds) Family Systems and Cultural Change, pp. 161–75. Oxford: Clarendon
Press.
Salvat, M., Vigo, M., Macbeth, H. and Bertranpetit, J. (1997) ‘Seasonality of Marriages
in Spanish and French Parishes in the Cerdanya Valley, Eastern Pyrenees’, Journal of
Biosocial Science 29(1): 51–62.
Schutz, A. and Luckmann, T. (1973) The Structures of the Life-World. Evanston, IL:
Northwestern University Press.
Shryock, H.S. and Seigal, J.S. (1976) The Methods and Materials of Demography. New
York: Academic Press.
Thang, N.M. and Swenson, I. (1996) ‘Variations in Vietnamese Marriages, Births and
Infant Deaths by Months of the Julian Calendar and Years of the Vietnamese and
Chinese Astrological Calendars’, Journal of Biosocial Science 28(3): 367–78.
Tobler, C. (1996) ‘Procreation Time: Pregnancy and Childbirth in EC Sex Equality
Law’, Time & Society 5(3): 363–83.
Trovato, F. and Odynak, D. (1993) ‘The Seasonality of Births in Canada and the
Provinces, 1881–1989: Theory and Analysis’, Canadian Studies in Population 20(1):
1–41.
Tuma, N.B. and Hannan, M. (1984) Social Dynamics: Models and Methods. San Diego,
CA: Academic Press.
United Nations (1996) The World’s Women 1995: Trends and Statistics. New York:
United Nations.
Van der Poel, H. (1997) ‘Leisure and the Modularization of Daily Life’, Time & Society
6(2/3): 171–94.
Werschler, T. and Halli, S. (1992) ‘The Seasonality of Births in Canada: A Comparison
with the Northern United States’, Population and Environment 14(1): 85–94.
Willekens, F.J. (1987) ‘The Marital Status Life Table’, in J. Bongaarts, T.K. Burch and
K.W. Wachter (eds) Family Demography: Methods and their Application, pp. 125–49.
Oxford: Clarendon Press.
Willekens, F. J. (1990) ‘Demographic Forecasting. State of the Art and Research Needs’,
in C.A. Hazeu and G.A.B. Frinking (eds) Emerging Issues in Demographic Research,
pp. 9–66. Amsterdam: Elsevier.
07_Mills 2/2/2000 2:55 pm Page 127
Willekens, F.J. (1997) ‘Duration Models for Single-decrement Life Tables (Elementary
Processes)’, unpublished manuscript, University of Groningen, The Netherlands.
Yi, Zeng (1985) ‘Marriage and Marriage Dissolution in China: A Marital Status Life
Table Analysis’, Netherlands Interuniversity Demographic Institute, working paper
no. 57.
Young, M. (1988) The Metronomic Society: Natural Rhythms and Human Timetables.
London: Thames & Hudson.
Young, M.L. (1992) ‘Analysing Household Histories’, in E. Berqúo and P. Xenos (eds)
Family Systems and Cultural Change, pp. 176–200. Oxford: Clarendon Press.
Zerubavel, E. (1981) Hidden Rhythms: Schedules and Calendars in Social Life.
Berkeley: University of California Press.
Zerubavel, E. (1982) ‘Easter and Passover: On Calendars and Group Identity’, American
Sociological Review 47(2): 284–9.
MELINDA MILLS is a PhD candidate in Demography at the University of

Groningen, The Netherlands. She obtained a graduate degree in Sociology
from the University of Alberta, Canada. Her main research interests include
time, the social theory of Giddens, life course, multistate and event history
methods. ADDRESS: Population Research Centre, Faculty of Spatial
Sciences, PO Box 800, University of Groningen, 9700 AV Groningen, The
Netherlands. [email: m.mills@frw.rug.nl]

El Tiempo en La Vida

Uploaded by

Copyright:

Available Formats

El Tiempo en La Vida

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

El Tiempo en La Vida

Uploaded by

Copyright:

Available Formats

07_Mills 2/2/2000 2:55 pm Page 91

Providing Space for Time

ABSTRACT. The article alerts those in the field of quantitative life

The objective of this article is to alert those using quantitative means of

92 TIME & SOCIETY 9(1)

assumption, he then proposed the epistemological approach of cohort analysis.

MILLS : PROVIDING SPACE FOR TIME 93

Previous life course studies

94 TIME & SOCIETY 9(1)

agency (psychological phenomena, action, choice), meso-level interaction

MILLS : PROVIDING SPACE FOR TIME 95

1. Different Forms of Temporality

96 TIME & SOCIETY 9(1)

‘symbolic time’, which is an abstract entity to summarize experiences and

Natural or cosmic time

MILLS : PROVIDING SPACE FOR TIME 97

increased sexual activity occurring at different latitudes due to variations in the

98 TIME & SOCIETY 9(1)

Biological and psychological human development

MILLS : PROVIDING SPACE FOR TIME 99

100 TIME & SOCIETY 9(1)

MILLS : PROVIDING SPACE FOR TIME 101

on the individual or family unit is often traced through the epistemological

102 TIME & SOCIETY 9(1)

Cultural and social time

MILLS : PROVIDING SPACE FOR TIME 103

former dimensions of space and time to embrace consciousness and experience,

104 TIME & SOCIETY 9(1)

relation to the family or production systems. The ontological effect of institu-

MILLS : PROVIDING SPACE FOR TIME 105

consideration is artificial peaks and drops attributed to the timing of official

106 TIME & SOCIETY 9(1)

MILLS : PROVIDING SPACE FOR TIME 107

MILLS : PROVIDING SPACE FOR TIME 109

the establishment of human contact and draw people together’ (Zerubavel,

2. Temporal Aspects of Data, Methods and Methodology

Data, methods and methodology are the epistemological instruments linking

110 TIME & SOCIETY 9(1)

retrospective. Prospective, or panel data, are collected at discrete points in time,

Timing of data collection: selectivity and age truncation

MILLS : PROVIDING SPACE FOR TIME 111

112 TIME & SOCIETY 9(1)

MILLS : PROVIDING SPACE FOR TIME 113

Source: National Family and Health Survey, 1992–3.

as over-representation of dates or ages ending in zero or five. This is attributed

114 TIME & SOCIETY 9(1)

MILLS : PROVIDING SPACE FOR TIME 115

Data gathering: understanding the ontological experience of temporality

116 TIME & SOCIETY 9(1)

problems in survey information. These events can be used as ‘milestones’ or

Self-registration: the influence of data collection on temporality

MILLS : PROVIDING SPACE FOR TIME 117

Methodology: the force of ontology on the epistemological approach

118 TIME & SOCIETY 9(1)

MILLS : PROVIDING SPACE FOR TIME 119

The past, the present and the future

120 TIME & SOCIETY 9(1)

Although there have been innovations to incorporate temporality in life course

MILLS : PROVIDING SPACE FOR TIME 121