Shuffling Softly, Sighing Deeply: A Digital Inquiry into Representations of Older Men and Women in Literature for Different Ages

Abstract

When gender is brought into concerns about older people, the emphasis often lies on stereotypes connected to older women, and few comparative studies have been conducted pertaining to the representation of the intersection between older age and gender in fiction. This article argues that not only children’s literature, traditionally considered to be a carrier of ideology, plays a large part in the target readership’s age socialization, but so do young adult and adult fiction. In a large corpus of 41 Dutch books written for different ages, the representation of older men and women is studied through the verbs, grammatical possessions and adjectives associated with the relevant fictional characters, which were extracted from the texts through the computational method of dependency parsing. Older adult characters featured most frequently in fiction for adults, where, more so than in the books for younger readers, they are depicted as being prone to illness, experiencing the effects of a deteriorating body and having a limited social network. In the books for children, little to no association between older adulthood and mortality was found in the data. Ageist stereotypes pertaining to both genders were found throughout the corpus. In terms of characterization, male older adults are associated more with physicality, including matters of illness and mobility, while character traits and emotions show up in a more varied manner in connection to female older characters.

1. Introduction

Recently, stereotypes connected to older women have been widely addressed in different media, from Jessica Bennett’s (2019) “I am (an Older) Woman. Hear Me Roar” in the New York Times to the categories such as “the coastal granny” and “Madonna” that Lieve Van De Velde (2022) identifies in De Standaard and Michele Hanson’s (2013) article in The Guardian that considers older women as the “glue that holds society together”. Older men are also subject to stereotyping, for example, when they are presented as wise mentors or grumpy seniors. However, these images are still more naturalized as their voices are less represented in the mass media. Furthermore, when gender is brought into concerns about older people, the emphasis often lies on “the growing numbers of older women”, which according to Jeff Hearn, holds true not only in political settings and popular media, but also in various fields of literature. His plea that “the sociology of aging […] should be fully, not partially, gendered” (Hearn 1995, p. 98) has evidently not entirely found a reflection in the mass media since he voiced it in 1995, as men are hardly seen berating the ageist representations they undergo based on the combination of their age and gender. By integrating age studies and children’s literature studies with the digital humanities, which applies digital tools to study the disciplines of the humanities, this article will gauge how older men and women are portrayed in fiction. Contemporary age studies depart from the premise that age is a social construct as much as gender; as the influential age critic Margaret M. Gullette puts it, “human beings are aged by culture” (Gullette 2004, p. 12).

Taking note of Hearn’s plea to study the relationship between gender and age more completely, the analysis will compare descriptions of older female and older male characters. Moreover, it will consider the differences between books intended for readers of different ages to obtain a better understanding of how the representation of older adults in children’s literature compares to adult literature. Children’s literature is one of the cultural discourses in which this construction of age takes place (Joosen 2018, p. 9); it carries ideologies and reflects, as well as helps to create, societal values, perpetuating descriptions of older characters and launching ideas about how they look, think, behave, and act. Children’s books thereby contribute to young readers’ age socialization (Stephens 1992; Hollindale 1988). This process is not completed at the end of childhood, however, as adolescents remain susceptible to negative images of older adulthood. Gullette states that “ageism may worsen for adolescents, perhaps from overhearing the sotto voce complaints or jokey-fearful allusions of their parents to ‘aging’” (Gullette 2017, p. 57). While Gullette refers to real-life interactions with adults, her observations open the door to the hypothesis that young adult literature continues to contribute to the socialization of its readers with regards to their perception of (older) age. In contrast to books for young readers, adult novels are targeted at a broad range of life stages, from early adulthood to people in deep old age. Thus, the representation of older adult characters in those books is consumed by people of, or close to, that same age. The representation of older age then becomes no longer merely a matter of socialization, but also one of self-reflection.

This article delves into the question of how adult authors construct senescence in children’s books, young adult novels and books for adults, while keeping in mind the intersection of age and gender. There are several interests in studying older age in fiction for different ages. First, as the average age of the worldwide population is rising, it is important that not only children, but also adolescents and adults, are “aware of age ideology” (Joosen 2021a, p. 182) so that they can also be critical of ageism. Second, Sandra McGuire (2016) argues that positive images of ageing are important for today’s children and how they age. While I agree with her statement that “the foundation for positive aging should begin during youth” (McGuire 2016, p. 2605), I believe it is important not to limit the study of the representation of older age solely to children’s literature as the process of developing imaginaries for positive aging continues throughout all life stages leading up to it. While holding a negative image of older age may seem beneficial to younger adults in a more immediate manner, such as paying lower insurance rates, in the long run it proves to be detrimental to their mental and physical health (Levy 2009). To put it another way: today’s children, adolescents and adults are influenced by their perception of older age not only in their interactions with older people in their daily lives, but also in the way they look towards their own futures, as they grow towards older age. Third, because of the points raised above, a better understanding of the portrayal of older adulthood in fiction is a necessary step towards equality based on age, whereas, at present, ageism, inequity and social injustice still abound.

In compiling a corpus of Dutch and Flemish fiction, I borrow the reasoning from two publications by children’s literature scholar Vanessa Joosen. First, Dutch literature has presented “a preoccupation with broadening the range of age narratives” over the last few decades (Joosen 2020, p. 4). Second, Dutch children’s books “display an original and progressive take on old age” in several ways, including “by highlighting the diversity in old age” (Joosen 2021a, p. 182). Rather than selecting literature written for readers of different ages by different authors, the texts in the corpus are all written by crosswriters. Bibi Dumon Tak, Ed Franck, Guus Kuijer, Bart Moeyaert, Hilde Vandermeeren and Joke van Leeuwen have all written novels for at least two categories of readers based on age (e.g., children’s literature and adult literature, or children’s literature and Young Adult literature). Celia Catlett Anderson argues that comparing passages by authors who address a different readership in different works is the best method to compare children’s and adult literature as it minimizes the differences according to the individual authors (Anderson 1984, p. 40). This article answers the following three research questions by studying the representation of older age in their work:

(1)

How do representations of older adulthood differ according to gender?

(2)

How do representations of older adulthood differ according to the age of the intended reader?

(3)

Can we find evidence of Dutch progressiveness by using a distant reading method for characterization?

This last question serves as an evaluation of the unique methodology this article employs to study older age on a large scale. The representation of fictional characters is studied in a quantitative manner, which relies on different types of descriptions based on the relevant adjectives, verbs and possessions. Studying these word types to better understand the representation of fictional characters is discussed by both the linguist Svenja Adolphs and the leading children’s literature researcher Maria Nikolajeva. The former claims that there are two fundamental methods of characterization; namely, “the verbs and adverbs that illustrate [a character’s] actions and in the adjectives that are used to describe the person” (Adolphs 2006, p. 67). In Nikolajeva’s narratological study The Rhetoric of Character (Nikolajeva 2002), she places different methods of characterization on a scale from direct to indirect methods, from description and narration to implicit characterization through a number of different constructs. In between these two extremes are the actions a character performs, the events they become embroiled in, their speech and their internal perceptions (Nikolajeva 2002, pp. 157–58). In this paper, adjectives represent the direct characterization on one end of that continuum, verbs represent the actions taken and possessives are studied as a type of implicit characterization. Combining these data that can be extracted from a large number of texts through the use of digital tools with insights from the field of age studies can improve our understanding of the construction of older age and how this life stage is represented in books for readers of different ages.

Since the last quarter of the twentieth century, several studies have been published that trace and criticize ageism and stereotypes in books for young people. However, some of these, including Marilyn Apseloff (1986) and Hilary Crew (2000), have only studied specific facets of older age, such as the representation of grandparent figures and daughter-mother narratives, respectively. Other studies, including Sylvia Henneberg (2010), situate themselves on the intersection of older age and gender, but often investigate only one gender, most often older female characters (see also Chivers 2003; Woodward 1999). Henneberg concludes that classic children’s stories perpetuate ageism and sexism. Since the mid-2010s, a higher awareness around adulthood and its construction in children’s literature has emerged. The first work to focus specifically on (older) adulthood as a life stage, as opposed to roles such as motherhood, fatherhood and grandparenthood, is Joosen’s Adulthood in Children’s Literature, in which she argues that age studies and children’s literature studies have many insights to offer each other, as they both acknowledge that age is a cultural construct as much as it is a biological given (Joosen 2018, p. 9). In this work, Joosen discusses four ageist tropes: the decline narrative, the infantilized senior, the disregard of the old body and the wise old mentor. The latter is an example of an ageist stereotype as a result of the inclusion of positive images of older adulthood. Age scholars argue that reproducing these types of images disregards the diversity present in older age.

While borrowing insights from the previously mentioned small-scaled studies, this article also builds upon the tradition of large-scale studies into the representation of older age in children’s books. From the very beginning of the studies mentioned above, there was an interest in seeking trends in a wide range of texts. In one of the first studies into older age in children’s literature, Phyllis Barnum (1977) conducted a close-reading analysis of 100 books for children from preschool through to grade three in the United States of America (USA). She found that “the aged play an insignificant role in young children’s literature” (Barnum 1977, p. 303) and that, when including gender into the analysis, older women are even less represented than older men when compared to the real-life distribution of older adults. Furthermore, older adult characters are portrayed as being excluded from society, as well as less healthy and less self-reliant than other adults (ibid., p. 304). In 1978, Edward Ansello published a paper entitled “Ageism—The Subtle Stereotype”, based on research of a corpus of 656 children’s books. Older adult characters featured in only 16% of these titles. Ansello (1978) studied the representation of these characters through close reading and found mainly stereotypical characterizations. The same conclusion was reached by Linda Janelli’s (1988) research into depictions of grandparents in 73 books for children in 1988.

In the last decade of the twentieth century, the stereotypical image of grandparents that researchers identified shifted. Beland and Mills (2001) studied 64 children’s books published between 1985 and 2000 and found that these older adult characters are often characterized as, among others, independent and happy, creating an overall positive portrayal. Ng et al.’s (2015) large-scale study of ageism moved towards a digital methodology, in which they identified an attitude opposite to that of Beland and Mills’ study. From collocations, words that co-occur frequently in a text, in a corpus of historical American-English of the past 200 years, Ng et al. found that age stereotypes have gradually become more negative, with a turning point in the 1880s, where they found that stereotypes switched from being positive to negative. While collocation is an accessible tool for extracting associated terms from texts, it has certain limits. For example, Ng et al. used a list of synonyms for the term ‘elderly’ as the starting point of their analysis, which potentially discounts characters who are not constantly referred to as elderly, senior or older. By using an annotated corpus as the data set, as well as more advanced techniques from the field of digital humanities, this article places itself firmly at the forefront of the search for new methods for studying older age in fiction.

2. Materials and Methods

The analyses in this article rely on tools from the digital humanities for two reasons. First, they make it feasible to examine specific aspects of a large corpus in an efficient way and to study different avenues of research in a relatively short time, bypassing the need to reread a large number of texts. A second advantage is that they can recognize patterns that are often less obvious in a traditional close reading, such as information about word frequency and the co-occurrence of words. The corpus analyzed below constitutes 41 contemporary books, published between 1975 and 2021 and written by Dutch and Flemish crosswriters. To study characterization, all of the texts were digitized and an additional layer of information pertaining to the fictional characters was added in a process of textual annotation. First, the books were either obtained in digital versions or physical copies were scanned and converted into editable text documents using Optical Character Recognition (OCR) software.

Table A1. List of novels included in the corpus of this article and the corresponding age of the intended reader.

Author	Title (Year of Publication)	Minimal Age of the Intended Reader
Bibi Dumon Tak	Mikis de ezeljongen (2011)	7
	Bezem (2005)	7
	Latino King (2010)	15
	Rotjongens (2007)	18 (book)
	De dag dat ik mijn naam veranderde (2020)	18 (book)
Ed Franck	Els heeft een pels (1995)	7
	Moet je echt weg? (1987)	7
	Spetters op de kermis (1985)	9
	Wie wil Wubbe weg? (1989)	9
	Begraaf me over de bergen (1988)	18
Guus Kuijer	Tin Toeval en de kunst van Madelief (1989)	7
	Ik ben Polleke, hoor! (2001)	9
	Met de wind mee naar zee (2001)	9
	Het land van de neushoornvogel (1985)	12
	Drie verschrikkelijke dagen (1976)	12
	Een gat in de grens (1975)	12
Bart Moeyaert	Die steeg van ons (1995)	7
	Mansoor (1996)	9
	Suzanne Dantine (1989)	12
	Terug naar af (1986)	12
	Broere (2000)	15
	Het is de liefde die we niet begrijpen (1999)	15
	Graz (2009)	18
Hilde Vandermeeren	Moord in de wijk (2019)	18 (book)
Joke van Leeuwen	Maar ik ben Frederik, zei Frederik (2013)	7
	Toen ik (2017)	7
	Ik heet Reinier en ons huis is afgebrand (2020)	9
	Slopie (2004)	9
	Toen mijn vader een struik werd (2010)	9
	Dit boek heet anders (1992)	9
	Het verhaal van Bobbel (1987)	9
	Iep! (1996)	9
	Kukel (1998)	9
	Wijd weg (1991)	9
	Bezoekjaren (1988)	12
	Alles nieuw (2008)	18 (website author)
	De onervarenen (2015)	18 (website author)
	De tjilpmachine (1990)	18 (book)
	Feest van het begin (2012)	18 (website author)
	Hier (2018)	18 (website author)
	Vrije vormen (2002)	18 (website author)

in Appendix A provides an overview of the works, along with the age of the intended reader for each, according to which the texts are categorized in the analyses in the next section. With only a few exceptions, most of the books in the corpus are recorded in the Centraal Bestand Kinderboeken (CBK), a Dutch database that includes information on the intended readers’ age. For titles not included in the catalogue, the table explicitly states the source of the reader’s age information; it was either taken from the physical book or the author’s website.

Second, annotations were added to various parts of the digitized texts. By adding tags to pronouns, proper names and nouns, all references to characters were disambiguated. These tags are formatted by using Extensive Mark-up Language (XML) in accordance with guidelines set up by the Text Encoding Initiative, a widely used standard in digital text analysis. The following fragment from Joke van Leeuwen’s Alles Nieuw illustrates the texts after annotation:

This house is a hundred years old, says the <rs ref=“ada80s”>landlady</rs>, who <rs ref=“laraearly”>I</rs> think is at least eighty. All predate <rs ref=“laraearly”>my</rs> time. This is <rs ref=“laraearly”>my</rs> time. <rs ref=“ada80s”>She</rs> lives downstairs, <rs ref=“laraearly”>I</rs> have to pay <rs ref=“ada80s”>her</rs> in cash every month, <rs ref=“ada80s”>she</rs> jots this down in a notebook, peering at the tip of <rs ref=“ada80s”>her</rs> pen, while <rs ref=“laraearly”>I</rs> sit looking at <rs ref=“ada80s”>her</rs> hands, at the map of veins and dark spots that could represent rivers and villages.1

Each reference to a character is enclosed by a so-called ‘rs-tag’, which is attributed with an id unique to each character. This excerpt contains two characters: Lara, who is also the narrator of the story, and Ada, her landlady. Their ids, ‘laraearly’ and ‘ada80s’, are recorded in a separate data frame (e.g., an Excel table), together with several character features, including name, gender, ethnicity and age. The latter is registered as either a numerical age or a life stage, as determined by the age scheme in

Table 1. Age scheme used for the annotations, categorizing numerical ages into categories of life stages.

Life Stage	Corresponding Numerical Age
infant	0–2
child	3–11
- early child	3–5
- middle child	6–8
- late child	9–11
adolescent	12–19
adult	20+
- early adult	20–39
- middle adult	40–59
- old adult	60–79
- deep old adult	80+

, which is drawn up for the annotations and is based on theories by age critics Lorraine Green (2010) and Thomas Armstrong (2007). Whenever the age of a character changes in the story, for example in a flashback or when the story spans several years, a new id is created to keep track of those changes. Where possible, the assignment of age relies on explicit mentions of the age or life stage of a character. According to Lara, Ada is “at least eighty” and her age is thus recorded as “eighties”, a life stage defined as spanning from 80 to 89. If there is no explicit mention of a character’s age in the text, age is derived from social markers and context. Later in Van Leeuwen’s story, the reader learns that Lara had moved out of her parents’ house six years earlier. Given that it is common in Western society for young adults to leave the parental home in late adolescence or early adulthood, and six years have passed, Lara was assigned the age ‘early adulthood’.

While context often provides the only indications of a character’s age, there are downsides to depending on social markers. The designation of Lara’s age relies on assumptions made about the age at which young people start living alone. The categories included in the age scheme are debatable and dependent upon social norms; it is also entirely possible that Lara only moved out of the parental home at a later age. The reliability of explicit mentions of a character’s age by another character can also be questioned. Fictional characters, as well as the human annotators of the texts, might rely on their own understanding of age to assign it to other characters. For example, Lara, who is identified as probably being a younger adult, might be wrong when estimating the age of people who are more than twice her age. Similarly, narrators can be unreliable due to their lack of omniscience, personal involvement in the story, or because of a deliberate deception or questionable value-scheme, in terms of characterization (Rimmon-Kenan 2002, pp. 103–4). These reservations have to be taken into account when relying on textual information to establish characters’ ages.

To assign characters to the life stages of ‘old adult’ and ‘deep old adult’, the categories that are of interest to the analyses in this paper, the annotation relied on explicit markers of age, namely a numerical age of at least 60 and variants of the word ‘old’, and social markers, such as retirement and family relations, including parents of adults and grandparents. Some parts of this method are again debatable. In real life, not all grandparents are older than 60 and so do not fit into the ‘old adult’ category. However, in children’s literature, most grandparent figures are portrayed as being older than 60 (Joosen, forthcoming). This age norm in children’s literature and in the annotation of the corpus for this article is self-perpetuating; as grandparents are mainly portrayed as older, all grandparent characters, except when explicitly assigned a younger age, are considered to be over the age of 60. Furthermore, relying on certain descriptions to assign the life stages of ‘old adult’ and ‘deep old adult’, such as the word ‘old’ or the physical feature of grey hair, can lead to circular reasoning when looking at the results; do older characters walk laboriously or are they labelled older because they walk laboriously? The interpretation of the results in the next section tries not to fall into this trap by taking the argumentations for the assignment of these categories into account.

After identifying all of the references to older adults in the 41 texts, all of the verbs and adjectives associated with, and words that stand in a grammatical possessive relation to (henceforth referred to simply as ‘possessions’), these characters are extracted to the representation of these life stages on a large scale. To this effect, a Python script was developed in collaboration with TEXTUA, the Text Mining Centre of the University of Antwerp.2 The script uses the open-source software spaCy (Honnibal and Montani 2017) to parse the syntactic dependencies in sentences. This results in a dependency tree, which links all of the words that are associated with each other, as well as the type of relationship they have; for example, a verb and its subject. By combining the resulting data from each text with the corresponding character list, the ‘parser’ can extract the relevant information of characters in different age groups.3 As the analyses in the next section seek to detect the differences between the representation of male and female older adult characters, group characters with mixed genders, e.g., ‘pensioners’, were removed from the data. After the extraction, all verbs, possessions and adjectives are lemmatized, a process which converts each word to its stem form, taking into account their part in speech. For example, depending on its intended meaning in the text, the term ‘bears’ would be interpreted as the plural form of the animal ‘bear’ or a conjugated form of the verb ‘to bear’. After this process, the fifty most frequently used verbs, possessives and adjectives are extracted. While these lists are a great source of data for studying the representation of older people in children’s literature, we need to take into account that a distorted picture may be presented. Extracting a single word removes possible negation and ignores nuance. To limit the bias that might arise from a naïve analysis of these words, the parser also extracts adverbs and adjectives pertaining to the verbs, possessions and adjectives associated with older age.

3. Results and Discussion

While the complete collection of annotated texts fed into the parser contains 97 titles, data were collected from only 41 of those. Sixty percent of the titles did not include sufficient data on older adult characters; if older adult characters feature in those stories, their role is insignificant or they are mentioned only fleetingly as none of them undertake an action, are attributed with an adjective or have a possession. When looking at the distribution of older adult characters across books for readers of different ages, books for adults contain the most older characters; data were extracted from 11 of the 16 titles in this category. One possible explanation is that part of the intended readership of these books are older people and the other part are in the stage of life that precedes older age. Including characters who share certain characteristics with the intended readership, in this case their age, increases the chances of the reader’s involvement with the story. The same happens in children’s literature, which usually features a large number of child characters. However, according to the most basic definitions of children’s literature, the ‘child’ in the term refers to the intended reader. Thus, while researchers, including Barbara Wall (1991) and Perry Nodelman (2008), argue that an adult almost always acts as a second reader or second addressee, these books are not initially written with (older) adults as the main target audience. Nonetheless, older adult characters do often play an important role in stories for children and traditionally have a close relationship with the child characters. In this corpus, however, the children’s books contained the smallest proportion of older adult characters; only around one-third of the titles for seven and nine year olds contained relevant information about this life stage. The close connection between children and older adults is generally lost when the child ages into adolescence (Joosen 2021a, pp. 184–85). While, from observation, one might derive that older characters will appear less frequently in books for young adults, just short of half of the books intended for young adults aged between twelve and fifteen feature older adult characters, a significantly larger percentage than the books for children.

In the next section, verbs, possessions and adjectives are discussed separately, according to the following structure: for each word type, several themes that emerged from the extracted word lists are analyzed, first in books for children, then in young adult novels and finally in adult novels. The themes that are included in the discussion are a selection of a handful of topics that were identified during a review of all six word lists for each word type, one for each gender over the three categories of books, according to the intended age of the reader. Some of these themes contain words that appear only in small numbers in the word lists but nonetheless contribute to the thematic cluster. The selection was made based on the relevance of a theme, as well as themes with the greatest number of words, with a preference for modified terms. A complete record of the word lists, including themes that are not discussed below, can be found in Appendix A, Appendix B, Appendix C and Appendix D.

3.1. Verbs

Extracting verbs that are syntactically dependent on older adult characters from texts for children, young adults and adults provides an insight into the different activities in which this life stage is associated in fiction. Most verbs could be assigned to one of the following topics: the physical and mental modes of perceiving and processing surroundings, verbs that describe movement, ways of expressing oneself, leisure activities and living in older age with illness and mortality (see Appendix B for a full word list). To keep the analysis of actions manageable and to the point, the following discussion covers all but the first of these themes. In the books for children aged between seven and nine, there is no large variety in the movements of older adult women; in terms of stationary actions, they ‘stand’, ‘sit’ and ‘stop’, while more dynamic verbs include ‘come’, ‘pull’, ‘shuffle’ and ‘walk’4. The latter is accompanied at least once by the adverbial phrase ‘no longer well’, which is very similar to one of the twenty most common words used by children aged between three and six to describe older people, as established by Allison Flamion et al.’s study, which included children aged between three and six and adults: “walks with difficulty” (Flamion et al. 2020, p. 7). This action, and more specifically its further description, suggests the stereotypical image of the slow older person walking with a laborious stride. Older male characters show a wider range of movement; they also ‘stand’ and ‘sit’, but their list of dynamic movements includes grander moves. Although they are not described as having more trouble walking, older men ‘fall’, which connects older age to frailty and pain. On the other hand, men ‘lift (something or someone) high’, which in turn indicates vitality in older age and contradicts the ageist image of the frail elder. Furthermore, older male characters ‘sneak’, ‘disappear’ and ‘hide well’, which may indicate a playful element. In children’s literature, older adults engage in play often when accompanied by a child; the fondness of play shared by grandparent and grandchild figures are a common trope (Joosen 2018, p. 189).

In the entire corpus, women in the books for children are the only older adult characters whose speech is not modified; they ‘say’, ‘talk’, ‘tell’ and ‘ask’, but the manner in which they do so is not specified. The lack of information on how these characters express themselves is no doubt due to the small proportion of older female characters who play an active role in the stories for children. The only exception is ‘strictly forbid’, which evokes the image of a strict and stern older woman. This stereotypical character image is further constructed by one of the emotions older women in the books for children are said to display; they ‘look angry’. However, this image is nuanced by the fact that they also ‘look grateful’ and ‘become cheerful’. Modified verbs pertaining to verbal and non-verbal expressions of older adult male characters also construct different types of characters. One of the depictions outlined is that of the obnoxious, angry older man who ‘says (something) gruffly and sternly’, ‘sighs deeply’, ‘laughs bitterly’ and ‘talks too loud’. Another character type emerges from the following modified verbs: ‘whisper incoherently and agitated’, ‘talk very softly’, ‘say sadly’ and ‘look sad and dreamy’. These descriptions evoke the image of a frail and downcast older person. A more cheerful picture is painted by the last group of verbs: ‘smile wide’, ‘put at ease’ and ‘look encouragingly and friendly’. A possible hint at the element of play, as discussed above, is also present in the word list of older adult male characters; they ‘say conspiratorially’.

Although ‘play with’ shows up in the twenty most common words in children’s descriptions of older adults (Flamion et al. 2020, p. 7), the references to play found in the actions of older adults in books for children remain largely implicit. Both genders do engage in ‘drawing’ and apparently do so ‘well’. Other leisurely activities include ‘baking’ for women and ‘cleaning’ for men, subverting the gender stereotype of women being in charge of housekeeping. Several actions undertaken by older adult characters in the books for children refer to life, death and the deterioration of health. Both men and women ‘die’, but also ‘exist’ and ‘live’. The latter is modified in the male word list by the adverbial phrase ‘for so long already’, which, together with ‘be too old and too skinny’, emphasizes the characters’ older age. In terms of health, older adult women ‘breathe (ever more) slowly’, while men ‘feel weak’.

The discussion above considered only the children’s books. In the books for young adults, the movement of older adult characters is presented in a slightly more stereotypical way than in the books for children. Apart from the static verbs, ‘stand’, ‘stop’ and ‘sit’, women ‘trip backwards’, ‘shuffle softly’ and it takes them a lot of effort to make certain moves; the verbs ‘stoop’, ‘hoist’ and ‘bend’ are all modified by the adverb ‘laboriously’. Furthermore, there are traces of an element of anger in the way older people move, as women ‘ram furiously’ and men ‘stamp frantically’. Many other actions related to the movement of male older adult characters emphasize the lack of it. Men ‘lie dead still’, ‘sit motionless’, ‘lean’ and ‘stand encouragingly’. When they do move, they ‘shuffle’, ‘walk’, ‘move’ and ‘turn around quickly’, the latter being the only modified verb that hints at vitality in male older age.

A more diverse picture is painted of older adult characters in young adult novels when looking at the ways in which they express themselves and the emotions they display. There is a lot of overlap in the neutral ways older men and women say something; both ‘ask’, ‘answer’, ‘tell’ and ‘say’ something. The modifiers extracted for the latter, however, differ. Women are said to express thoughtfulness, urgency and kindness as they speak, while men are concise and despondent. Another difference found between the word lists of female and male older adult characters is the variety of verbs associated with both genders to express happiness and enjoyment, on one hand, and sadness and distress on the other. For the former, female characters only ‘laugh’ and ‘giggle’, while men ‘smile’ and ‘chuckle’ in addition to those. By contrast, as women ‘hit’, ‘shout’, ‘wail’ and ‘cry’, only the first two of these verbs are extracted for men. This suggests that older adult male characters in young adult novels are more often represented as being happy and older women are more often found in a state of distress, which is reminiscent of the stereotype of the hysterical hag, as identified by Susan Pickard (2016). The adverbs accompanying the verb ‘look’ supports this hypothesis; women ‘look anxious’, while men ‘look elated and serene’.

In addition to the references to more laborious movement, as discussed above, there are few references to illness in older age in books for young adults. In contrast to books for children, older people in these stories do not feel ‘too old’ or ‘weak’. The only exception is that older women ‘breathe excessively’ and men ‘leave (something) behind’, which may indicate the making of a will due to a higher awareness of mortality traditionally attributed to people of advanced age. Leisurely activities, on the other hand, receive more attention in young adult novels, but are exclusively associated with male characters. They ‘plant’, ‘build’, ‘drive’ and ‘knead’, which are actions that fit in with the stereotypical activities that Patricia Crawford lists for grandfather figures in children’s books, namely “taking walks or fishing” (Crawford 2000, p. 162).

In the adult novels in the corpus, older adult characters are represented as engaging more often in communicating and expressing their emotions than in active movement. When they are active, older characters are shown interacting with objects more often than in the books for young readers. This can be inferred from the high number of verbs of object movement in the fifty most common for both men and women. More so for older men than for older female characters, adverbs emphasize laborious movements. Men ‘push’, ‘pull close’, ‘slide far and open’, ‘throw hard’ and ‘stroke rhythmically’, but they also ‘shift’ and ‘drag’, both ‘laboriously’. In the list of verbs associated with female characters, there is no indication that they have difficulty moving objects; they ‘pull’, ‘push open wider’, ‘shake vigorously’ and ‘close well’. As in the books for younger readers, there are also verbs that refer to static and dynamic postures or movements of older adults, including ‘stand’, ‘sit’ and ‘lie’, as well as ‘walk’ and ‘step’, respectively. However, there is less diversity in these types of actions in the adult novels. A second difference is that, with the exception of ‘walk slowly’ for older men and ‘step carefully’ for women, there is no reference to a decrease in mobility.

Whereas the books for young readers relate more verbs referring to movement with older age, a large proportion of the fifty most common verbs in the adult titles refer to expression and emotion. The most variety is found in the word list of older female characters, with both verbal and non-verbal expressions, as well as a large range of emotions (see Appendix B for full list). To select a few, women ‘say’ things ‘softly’, ‘angrily’ and ‘cheerfully’, ‘speak very articulately’ and ‘greet briefly’. A number of other verbs hint at a negative image of older age; women ‘call desperately’, ‘shout anxiously’ and ‘helplessly’, ‘feel lonely’, ‘miserable’ and ‘restless’ and ‘become uncomfortable’. Furthermore, some adverbs evoke the ageist image of the rambling granny: ‘tell imperturbably’, ‘long-windedly’ and ‘talk incessantly’. Male older characters are depicted as less communicative in books for adults and, together with the emotions they display, the image of an angry older man appears. Older men ‘say directly’ and ‘conspiratorially’, ‘interfere’, ‘shout’, ‘cry’, ‘become mad’ and ‘grumpy’ and ‘look suspicious’.

Adults, more so than children, associate older age with “disease”, “health”, “dependence” and “frail[ty]” (Flamion et al. 2020, p. 7). A parallel observation can be made for the representation of matters of life, death and health in the adult novels in the corpus when compared to the books for younger readers. While in the latter, references to illness are made rather subtly (their breath comes ‘ever more slowly’ or ‘excessively’), the former explicitly describes older adult characters as ‘feeling sick’ and ‘looking sick’, both of which are modified verbs found in the female word list. Furthermore, the advanced age of both genders is suddenly seen as unduly high as they ‘become too old’, ‘live too long’ and ‘feel ancient’, as opposed to just ‘long’ or ‘old’. It is perhaps not surprising then that these ‘too old people’, who have already been shown not to have a wide range of motion in books for adults, are not often depicted performing a leisurely activity. Moreover, the few activities older women engage in paint a stereotypical picture of their occupations; they ‘polish shinily’ and ‘scrub too hard’.

3.2. Possessions

The possessions of older adult characters that the parser extracted from the corpus mainly relate to the following topics: body parts, health, personal relationships and community, home and household objects and clothing. The discussion in this section will focus on the first three topics (see Appendix C for wordlists of all five topics). In the books for readers aged between seven and nine, for all of these themes, there is more variety in the possessions of male characters and they are modified more often. The only female body parts that are described are ‘head’ and ‘hair’, which are, respectively, ‘small’ and ‘long’. The latter stands in opposition to the stereotypical image of older women with short hairstyles, being short hair or a bun, found by Villar and Fabà (2012) to be prevalent among primary school children. What is strikingly missing as a modifier of ‘hair’ is its color; ‘white’ or ‘grey’. This familiar image was found by Flamion et al. (2020) as one of the twenty most common words children use when describing older people, but only appears in this corpus in the adult novels. It is possible that this information is omitted from the text of children’s books because it is clearly visible in the illustrations included in or on the books. More so than for females, male body parts extracted from the books for children point toward an emphasis on the deteriorating body in older age. Older men are described as having a ‘skinny body’, ‘skinny hands’, ‘old mouth’ and ‘old tongue’, ‘bony fingers’, ‘unwashed skin’ and ‘wrinkles’. The same theme of deterioration, again mainly present in male possessions, is found in words associated with health and attributes used to improve quality of life. While ‘life’ is included in possessions of both genders (‘whole life’ in the male list), the more abstract ‘existence’ and the reference to mortality of ‘grave’ feature only in those of older male characters. Both older women and men seem to wear ‘glasses’, but only the latter require ‘dentures’ to live their lives comfortably. The encompassing word ‘health’ also appears only as a possession of men, indicating that this is not only a theme for those characters, but is also thematized in the children’s novels.

Looking at grammatically possessive relationships, family members account for the largest proportion of people in the communities of older characters in the books for children. This is mainly the case among men, while the relationships of older women appear to be more varied. In the list of women’s possessions, the following people outside the family sphere are named: neighbor, locksmith, student and class. The latter two suggest an educational work environment, contradicting the image of the retired older person who disengages from professional communities. This stereotypical image fits within disengagement theory, constructed in 1961 as part of social gerontology, which states that it is the norm for older people to withdraw from social relationships and suffer a decrease in societal roles (Cumming and Henry 1961). Although this theory was soon criticized, partly because it encourages ageism and partly because the statements are not supported by empirical research, this negative view of older adults often creeps into the way this life stage is considered in contemporary society. In 1990, Fraboni, Saltstone and Hughes developed the Fraboni Scale of Ageism, a method to measure this attitude towards older people. One of the items on which their scale is based is the following: “Many old people are not interested in making new friends, preferring instead the circle of friends they have had for years” (Fraboni et al. 1990, p. 62). Margaret Morganroth Gullette argues that generalizations such as this are largely accepted by young people (Gullette 2017, p. 58). The books for children aged between seven and nine studied in this article seem unaffected by this aspect of ageism, as in the list of possessions associated with older male characters, the only person mentioned who is not a relative is a ‘new friend’, breaking the image of the disengaging older adult. This observation was also made by Joosen (2015), who analyzed four children’s books from the Netherlands and the UK in search of the representation of older people according to four ageist tropes.

In parallel to the books for children, the young adult novels included in the corpus (intended for twelve-to-fifteen year olds) include more data on older male than female characters, albeit with a smaller distinction. Most of the body parts named of older characters do not immediately evoke stereotypical images. With the exception perhaps of ‘lap’ and ‘joint’ (both from the male list), which can be argued to evoke a predominantly older image, words such as ‘shoulder’ and ‘rib’ can be associated with any age. However, when we look at the adjectives used to describe body parts, patterns emerge for both genders. On the one hand, older men have ‘thin hair’, a ‘wiry body’ and ‘bony hands’. The latter is also included in the female list, together with ‘dry hands’ and ‘last breath’. On the other hand, older women in young adult novels can also have ‘fleshy arms’ and ‘chubby fingers’, suggesting possible obesity. Although this interpretation makes it seem as though older people are depicted as being either under- or over-weight, this contrast is most likely created by the nature of this study; by mainly focusing the discussion on possessions attributed with an adjective, references such as ‘arms’, ‘legs’, etc., without an adjective, which are also present in the data, fade into the background.

No adjectives are included for the possessions pertaining the health of older adult characters in the books for young adults. Based on gender, two distinct images of older people occur; the female list contains the words ‘death’ and ‘deathbed’, while the male list features the more positive words ‘movement’ and ‘wheelchair’. The latter emphasizes the attributes that older people have at their disposal to stay mobile. In Flamion et al.’s (2020) study, three of the twenty most frequently used words to describe older people pertained to assistive devices. Adults barely, if at all, referred to such attributes. In this regard, the construction of older age in the books for young adults in this corpus is more closely related to children’s perceptions compared to those of adults. In terms of the community around older adult characters, there is relatively less variety in the young adult books than in the titles for children discussed above. The list of female relationships is short (‘man’, ‘daughter’ and ‘friend’), while that of male characters is slightly more varied. In addition to predominantly family members, there are also references to older men’s ‘former people’ and ‘first customer’. The latter is reminiscent of the ‘new friend’ in the books for seven-to-nine year olds that indicates a wider social circle than just family.

In the adult novels included in the corpus, there is less difference in the proportion of data from men and from women, and possessions are described more often than in the books for younger readers. The characterization that emerges from these words is predominantly stereotypical. Although few body parts are attributed with an adjective in both the male and female lists, older age is similarly represented in both genders. Older women have ‘dry eyes’, a ‘worn-out’ and ‘rigid body’, an ‘uncooperative’ and ‘stiff back’, ‘old fingers’ and ‘grey’, ‘tangled’, ‘thin white’ and ‘uncombed hair’. Men are described as having ‘blotchy hands’, ‘cold cheeks’, ‘frail’ and ‘weak legs’, a ‘hairy’ and ‘hunched back’, ‘yellowish whites of the eyes’ and a ‘heavy’, ‘unwieldy’ and ‘slow body’. This emphasis on the deteriorating older body is reflected in the adult descriptions of older people in Flamion et al.’s (2020) aforementioned study, where ‘frail’, ‘disease’, ‘death’ and ‘health’ can be found among the twenty most common words. On the other hand, however, positive words such as ‘experience’, ‘wisdom’, ‘kindness’ and ‘love’ can also be found in their descriptions, which do not feature in the lists of possessions extracted from the adult books in the corpus. The words associated with health and the associated devices paint an equally bleak picture; older female characters wear ‘thick glasses’, but beyond that only references to ‘death’ and ‘deathbed’ are made. The former also features in the male list, together with ‘birthday’.

From what we can derive from the possessives, the community around older people in books for adults is significantly limited and almost exclusively includes close family members. The only exception is found in the male list, which includes ‘murderer’, linking older age to mortality. Furthermore, the only mention to a dead relative occurs here, as one older adult male character is described as having a ‘dead wife’. Novels for adults seem to be more explicit in their connection of ageing and death, as “children’s books are quite selective in the topics they address” (Joosen 2018, p. 24). A further gap that can be identified in the texts for young readers in this corpus and is filled in the adult novels is embodied in the ‘absent daughter’ as a possession of older female characters. In her discussion of senior parents and their adult children, Ellyn Lem states that caring for an ageing parent can be a complex responsibility and carries the risk of leading to estrangement (Lem 2020, p. 30). The resulting loneliness suffered by the parent “is often not communicated due to a lingering stigma attached to it” (ibid., p. 42). While absent adult figures, often parental, are a common trope in children’s books, for the benefit of the child-grandparent relationship, their absence is not usually labelled as such. The systematic equation of older adulthood and loneliness and the deterioration of their social relations, as discussed above, is confirmed by Flamion et al., where the most frequently occurring word with which adults describe older people is ‘loneliness’ (Flamion et al. 2020, p. 7). Standing close to older adulthood as their next phase in life, being abandoned by their own children seems to be a fear that adult authors are more likely to express in books for adults than for children.

3.3. Adjectives

Although the analysis of adjectives probably comes to mind first when wanting to study characterization in fiction, this category contains the least amount of data. In her monograph on characterization in children’s books, Nikolajeva identifies gender stereotyping in the direct description of characters, where mainly female characters are described, as opposed to men (Nikolajeva 2002, p. 189). However, this is not reflected in the corpus of this article. Mainly in the children’s books and books for young adults, there is more variation in the adjectives associated with men. This may be due to a possible smaller number of female older adult characters that play a role in those books. Across the genders, most of the adjectives associated with older adults pertain to one of the following categories: appearance, emotions, character traits, illness and age (see Appendix D for the full word list). In the discussion below, the first three will be examined in more detail. The latter is not included in the analysis, as both genders in all three groups of texts (children’s literature, young adult novels and adult novels) are described mainly as ‘old’. It is very likely that this description is an identifier rather than a description; during the annotation of the texts, age is assigned to characters based on explicit age markers, of which ‘old’ is the most obvious one for older adulthood. Characters are thus categorized as older people by the presence of the adjective ‘old’, rather than the adjective being a feature of characters otherwise considered to be older adults.

In the books for children aged seven to nine, female older adult characters are stereotypically described only as ‘grey’, while men are ‘dusty’, ‘small’ and both ‘fat’ and ‘skinny’. Apart from looking angry and grateful, which was discussed in the section on the actions of older women, no emotions are attributed to these characters in the books for children. Older men display a wider range of emotions, both positive and negative; they are ‘not scared’ and ‘maybe happy’, as well as ‘sad’. Furthermore, men are often depicted as being bitter characters in the children’s books in the corpus. The adjective ‘angry’ is included in the word list several times and is further described with the modifiers ‘maybe’, ‘very’ and ‘also’. In terms of character traits, this characterization emerges only from the adjective ‘strict’. Other direct descriptions of older men are much more positive: ‘kind’, ‘well-behaved’, ‘clever’, ‘valuable’. It is also this picture of older adulthood that is painted from the description of children, as observed by Flamion et al. (2020). References to negative emotions do not appear in the twenty most frequent words of the participants in that research, but ‘kind’ does (ibid., p. 7). Again, there is less data for older female characters; their character is described as ‘patient’, ‘perhaps more interesting’ and ‘thrifty’.

In the books for young adults, the emphasis in the direct descriptions of older adult characters lies mainly on their appearance and, for women, on their personality. From the former, the stereotypical image of the deteriorating body of the aged person emerges for both genders. Women are ‘small’, ‘grey’ and ‘wrinkly’, while men are ‘skinny’, ‘white’ and ‘thin’. One male character is also described as ‘unwashed’, which may indicate bad care. The same character is described with the only emotion associated with older male characters in the books for young adults, ‘dissatisfied’, which stands in contrast to the emotions of serenity and elation assigned to these characters in the analysis of verbs. Similarly, for older women in the young adult novels, only one adjective referring to an emotion was extracted from the corpus, which also goes against their display of anxiety previously discussed; they are depicted more positively, as being ‘happy’. For women, more data were extracted pertaining to character traits, but they are represented in a similar vein, as being ‘tough’, ‘always kind’ and ‘honest’. The single character trait mentioned as an adjective for older adult men in the young adult books is ‘secretive’.

From the adult novels, roughly the same number of total adjectives were extracted for both genders. In terms of appearance and emotions, the description of female older adult characters is more diverse, while men are attributed with a wider array of character traits. None of the adjectives describing the appearance of older people refer to the stereotypical image of the older body as being in a state of deterioration. Women are ‘flat’, ‘already a bit less blotchy’, ‘tall’ and ‘sweaty’. Men are also described as ‘tall’, as well as ‘heavy’. In terms of emotions, the list of extracted words associated with older women contains contrasting adjectives, painting a diverse picture of these characters. They are both ‘always happy’ and ‘not happy’, ‘very upset’ and ‘very relieved’, ‘grateful’ and ‘desperate’. The list of men’s emotions is shorter, containing only ‘happy’, ‘angry’ and ‘proud’. Finally, the character traits of older adult characters in the adult literature reflect an ageist trope for each gender. In her study of the history and meaning of fairy tales, Marina Warner identifies the prying older woman as a stereotypical character that has long since existed in literature (Warner 1994, pp. 27–50). Older female characters being described solely as ‘very curious’ in the adult novels might allude to this image. Furthermore, in addition to being ‘interesting’, ‘good’ and ‘not climate neutral’, older men are associated with the character trait of wisdom. Taken together with the lack of direct descriptions of older men referring to declining health, both physical and mental, this association seems to exemplify the ageist archetype of the wise older man as a denial of the decline narrative. This carries forth “the idea that with old age comes wisdom and that the old still have their knowledge and experience to offer to younger generations” (Joosen 2018, p. 194). Flamion et al.’s (2020) study confirms that this stereotypical equation of older age and wisdom is also made by adults in real-life situations. Among the top twenty words used by adults to describe older people are both ‘wisdom’ and ‘knowledge’ (Flamion et al. 2020, p. 7).

4. Conclusions

Several conclusions can be drawn from taking the observations on verbs, possessions and adjectives together. With regard to gender differences, older women’s mobility in the corpus of Dutch texts is underrepresented when compared to that of older male characters, but the laborious movements described for the latter category both confirm and contradict physical deterioration. Words referring to illness are associated more often with the male figures, but they are generally also attributed with a larger variety of assistive devices, suggesting that the effects of a deteriorating body can, in part, be overcome. Contrastingly, the range of expressions, character traits and emotions attributed to older men is less diverse than that of women in this corpus, portraying them as flatter and less communicative characters. A predominant emotion among men seems to be anger; the stereotypical image of the angry older man surfaced several times in the analyses. Other stereotypes identified in the corpus are the wise older man, the nosy older lady and the distressed, rambling granny. In contrast to this image of the unpleasant older man, many more elements that can be associated with play emerge for this gender, facilitating the connection between male older age and childhood.

Looking at differences between texts written for readers of different ages also reveals certain patterns, but maybe more significantly, several gaps. While age is considered as a significant characteristic of older adults equally in the books for children, young adults and adults, the treatment of this characteristic differs. In the former, an emphasis is placed on the length of the lives of older adults, but no association with mortality emerges from the verbs, adjectives and possessions associated with older adults in the corpus. This shifts in the books for adolescents, where references to death are more common, for example in the verb ‘leave behind’. In the adult novels in the corpus, the age of older people is sometimes seen as unduly high, as they ‘become too old’ and ‘live too long’, solidifying the connection between ageing and dying. The adult novels also portray older adulthood as a stage in life where people are prone to illness and experience the effects of a deteriorating body. While this is a realistic, if perhaps exaggerated, representation, older people in the books for children are less often depicted that way, pointing to a gap in the books for younger readers in the corpus. The image of the deteriorating older body is, however, included in the young adult novels. Moreover, the analysis suggests that older characters in the adult novels have few connections with people outside their household and almost none with people outside their close familial circle, constructing a sense of loneliness.

While the dense summary given above attests to the benefits of applying digital tools to the study of the construction of older age in literature for readers of different ages, the method did not pick up on the progressive themes that Dutch children’s literature is known for—this can be either because the books do not display this trend or because the themes are expressed in ways that the method does not include (e.g., illustrations, metaphors, more subtle suggestion). Based on the occurrence of verbs, possessions and adjectives, it seems that illness and death featured less prominently in the books for children than those for older readers, as did loneliness in older age. Finally, the analysis did not take into account the elements of direct speech and focalization. All of the descriptions of older characters were treated equally and not called into question, although, in a story, it can make a difference whether an ageist remark is made by a likeable character or by an antagonist. Although the texts in the corpus have been annotated not only to identify references to characters but also to attribute speech instances to characters, the scope of this article did not allow for including this extra layer in the analyses. However, an analysis focusing on only one category of texts, for example adult novels, could incorporate this element.

In a discussion of the significance of positive images of ageing in early children’s literature, McGuire cites that it is important for educators to first “evaluate one’s own attitudes about aging. Ageist attitudes are contagious and children will assimilate the attitudes of those around them” (McGuire 2016, p. 2605). I argue, with this article, that this susceptibility to a negative attitude towards older age does not stop when childhood is over. The promotion of positive aging, as important as it is for children and their interaction with older people in their daily lives, is equally important for adolescents and adults at different stages. On one hand, adolescents, in their search for identity, are more so than children confronted with the fact that soon they will be part of ‘the adult world’ in all its facets and that they, too, will one day become older. On the other hand, most adults will, at a certain point, suddenly have to carry the responsibility of caring for their own parents, as well as thinking increasingly about their own older age. The consequences of negative perceptions of older age among adults are abundant. To name just two, adults might develop a fear or unwillingness to age, as well as exclude older people from situations that are about them. For example, Joosen noted that, in 2020, it took months for an older adult to appear in a Flemish current affairs program about the recent corona pandemic, when they were the ones hit hardest (Joosen 2022, p. 6). It is also noteworthy that, while panels in such programs and in political debates, which are often organized by younger adults, increasingly pay attention to a proper representation of men and women, supplemented by spokespeople for minority groups, they further fail to take into account an equal distribution of the age of those participants.

Large-scale studies into the representation of older adulthood in fiction have an important role to play in shifting Western thinking towards a more diverse representation of older age, which is an important step towards active aging, a paradigm that recently emerged in gerontology that considers aging from a positive perspective (Fernández-Ballesteros et al. 2013). It is, however, important that this movement does not only take diversity on an individual level into account, but it also should not neglect the advances with regards to gender inclusion. The analyses documented in this article are the first to use tools and methods from the field of digital humanities to investigate the representation of female and male older adult characters in children’s literature, young adult fiction and adult novels in a comparative study. Due to the combination of a fairly large number of texts, namely 41, studied simultaneously through a fine-grained approach of digitally extracting words associated to characters in a specific life stage, patterns and gaps could be identified in the comparison based on gender, as well as in the different categories of texts according to the age of the intended reader.

The fact that ageist representations of older adults seem to appear predominantly in the adult books in the corpus calls into question why, in the effort to eradicate such representations, there are not more studies into the traces of ageism in these books. While the arguments that children assimilate ageism at an early age (Montepare and Zebrowitz 2002) are valid and the call for research in children’s literature remains important, I think there is still much to learn from studying books for young adults and adults. To carry out further research on a large scale, the methodology of this article is a useful tool for studying certain facets of characterization, especially if it is matched with analysis through close reading, which helps to contextualize the results for specific titles. By enabling the extraction of a lot of data, the method can help researchers to identify trends and select themes to explore further, taking us closer to diversifying the images of older adults in literature.