Declaration of authenticity
MA Applied Linguistics - 2016/2017MA-thesis Student name: Justine Griffioen
Student number: S2986329
PLAGIARISM is the presentation by a student of an assignment or piece of work which has in fact been copied in whole, in part, or in paraphrase from another student's work, or from any other source (e.g. published books or periodicals or material from Internet sites), without due acknowledgement in the text.
TEAMWORK: Students are encouraged to work with each other to develop their generic skills and increase their knowledge and understanding of the curriculum. Such teamwork includes general discussion and sharing of ideas on the curriculum. All written work must however (without specific authorization to the contrary) be done by individual students. Students are neither permitted to copy any part of another student’s work nor permitted to allow their own work to be copied by other students.
DECLARATION
• I declare that all work submitted for assessment of this MA-thesis is my own work and does not involve plagiarism or teamwork other than that authorised in the general terms above or that authorized and documented for any particular piece of work.
Signed
Integrating Perceptions of Aging in the
Third-Age Language Learning Classroom
J.C.M. Griffioen
Word count: 17,408
MA thesis
Applied Linguistics
Faculty of Arts
University of Groningen
Supervisor:
Prof. Dr. M.C.J. Keijzer
Second assessor:
Prof. Dr. Wander Lowie
2 Thesis outline
Abstract ... 3
Introduction ... 3
The Current Study ... 10
Research Questions ... 10
Hypotheses ... 11
Part 1: State of the Art ... 12
Defining the Third Age ... 12
The University of the Third Age ... 14
Literature Review (Effects) of Third-Age Language Learning (TALL) ... 18
The Socio-affective Effects of TALL ... 19
The Cognitive Effects of TALL ... 22
The Linguistic Effects of TALL ... 31
Stereotyping of Learning in the Third Age ... 35
Perceptions of Aging in the TALL Classroom ... 37
Conclusion Part 1 ... 40
Part 2: Research Agenda and Proposal... 43
The Terminology of Third-Age Language Learning ... 44
Focus Groups ... 45
Concept Mapping ... 45
Perceptions of Aging in TALL Methodologies ... 48
Older vs Younger Language Learner Competence ... 48
Individual Language Learner competence ... 49
Future Research Avenues ... 51
Positive Psychology as a Metatheory for TALL Research ... 52
Conclusion Part 2 ... 54
General Discussion ... 56
3 Abstract
Third-age language learning (TALL) is an emerging field of study. Due to the increasing aging population in especially (but not exclusively) Western parts of the world, age-related illnesses have started to become an increasing priority for health care systems all around the world (Gabryś-Barker, 2017). The third age, most strictly defined as the period of time in which seniors are retired, but still in good physical health (Moen, 2011), provides a window of time in which the onset of age-related illnesses could potentially be postponed. Second language learning could be one of the ways to achieve this. Indeed, researchers have attempted to chart the life experiences that lead to cognitive reserve in old age and, thus, healthy aging (Pfenninger & Singleton, 2019). Bilingualism has been claimed to be one of those life experiences. Moreover, bilingualism has been found to be a contributing factor in delaying the onset of dementia in old age by a number of years (Alladi et al., 2013). At the same time, researchers have attempted to examine which cognitive training programs can enhance cognitive flexibility in older adults with the aim of delaying the onset of most notably neurodegenerative diseases (cf. Park et al., 2014, among others). Given what is now known about the effects of lifelong bilingualism, an important question becomes whether language learning in older adulthood can itself function as a cognitive training regime. This thesis consists of two parts. Part one is a state-of-the-art account of the research regarding the cognitive, socio-affective and linguistic effects of learning a second language during the third-age. Furthermore, it addresses the issue of ageist stereotypes of learning in general in old-age. Part two is a research agenda and proposal, which aims to provide ways to improve stereotyping of learning in older adulthood with regard to terminology and methodology and proposes a metatheory for future TALL research.
Introduction
With an increasingly large portion of citizens in especially the Western world reaching an age of 65 or older, it is important to adopt strategies to combat third age-related ailments such as cognitive decline, depression and loneliness (Gabryś-Barker, 2017). It has been suggested that cognitive reserve most notably can serve as a safeguard against some of these old-age disorders (Whalley et al., 2004). Cognitive reserve is “the hypothesized capacity of the mature adult brain to sustain the effects of disease or injury sufficient to cause clinical dementia in an individual possessing less cognitive reserve” (Whalley et al., 2004, p. 374). Researchers have attempted to chart the life experiences that lead to cognitive reserve in old age and, thus, contribute to healthy aging (Pfenninger & Singleton, 2019). Bilingualism has been claimed to
4 be one of those life experiences. Strikingly, it has been found to be a contributing factor in delaying the onset of dementia in old age by a number of years (Alladi et al., 2013). At the same time, researchers have attempted to examine which cognitive training programs can enhance cognitive flexibility in older adults with the aim of delaying the onset of most notably neurodegenerative diseases (cf. Park et al., 2014, among others). Given what is now known about lifelong bilingualism as such an attenuating factor, an important question becomes whether language learning in older adulthood can itself function as a cognitive training regime. In other words, can the learning of a new language in older adulthood approximate effects that we witness as a result of lifelong bilingualism? Even though it is often argued that cognitive abilities decline with age, research has suggested that learning can still take place during advanced age (Ramírez Gómez, 2016; Ramscar, Hendrix, Shaoul, Milin & Baayen, 2014).
The perspective that learning in older adulthood is still a viable possibility is also reflected in other approaches that perceive aging in a different way than along the lines of decline. According to Ramscar et al. (2014), for instance, healthy seniors do not show cognitive decline when it comes to information processing: “Psychometric tests do not take account of the statistical skew of human experience, or the way knowledge increases with experience. As a consequence, when these tests are used to compare age groups, they paint a misleading picture of cognitive development” (Ramscar, 2014, p. 7). Specifically in relation to language processing, older adults have consistently been reported to take longer in searching for words (Verhaegen & Poncelet, 2013). From Ramscar’s perspective, vocabulary grows with experience over the course of a lifetime, and the time needed to sort through our memory thus becomes longer by default, and therefore leads to increased difficulty with information processing. Rather than labelling this process as cognitive decline, Ramscar claims, it is more accurately described as the result of accumulated knowledge that takes longer to sort through.
5 In spite of these findings, narratives of decline still tend to obstruct the lifelong learning paradigm as they are engrained in society’s views and attitudes towards old age. Very revealing in this respect is a study looking into a corpus of 400 million words taken from the Corpus of Historical American English (COHA) (Ng, Allore, Trentalange, Monin & Levy, 2015). This database includes around 100,000 texts which were published between 1810 and 2009 and come from a diverse range of written sources (from magazines to books). On the basis of this corpus, Ng et al. (2015) found that age stereotypes have become increasingly negative over the past 200 years due to the increased medicalization of elderly people and the societal burden associated with the aging population. For the study, a list of synonyms for the word ‘elderly’ was compiled on the basis of the corpus, for the period from 1810 until 2009, in order to track the different associations of the word over time. Following this procedure, 100 collocates were identified. They were included in the study if three inclusion criteria were met: (1) they occurred within a range of four words to the ‘elderly’ synonym (lexical item) in order to account for lexical proximity; (2) they related or referred to an older adult in particular; and (3) they more strongly associated with the elderly synonym than other words that occurred in the corpus for that decade. The collocates were placed on a spectrum (on a scale from 1 to 5) in order to indicate the level of negativity of the stereotype (the age-stereotype score) per decade, with 5 being the most negative. In addition, each collocate was rated as related or unrelated to physical health or illness in order to examine the medicalization of older people. After this, both scores were summed for each elderly synonym and weighted on the basis of how frequently they occurred in that decade. It was found that, from 1880 onwards, stereotypes have become more and more negative in connotation, in contrast to the preceding period from 1810-1880, where most age stereotypes were found to be positive. In addition, medicalization of aging was found to significantly predict negative age stereotypes, as every 1% increase in collocates related to the medicalization of aging was associated with a 20.5% increase in negative age stereotypes
6 over 200 years (Ng et al., 2015). It should, however, be noted that the life expectancy of people in the 19th century was lower and societies would therefore be less ‘burdened’ with old age illnesses in general and medical interventions did not take place on the same scale as nowadays.
Such negative age stereotypes have permeated societal views of what seniors can still learn in older adulthood. In a cross-cultural study by Löckenhoff et al. (2009), 3435 college students from 26 cultures divided over six continents completed questionnaires related to perceptions of aging. One of the scales which was used looked at the following factors (amongst others): 1) the ability to do everyday tasks; and 2) the ability to learn new information; 3) general knowledge; and 4) wisdom. Participants then had to respond to statements on a Likert scale ranging from -2 (this characteristic decreases a lot with aging) to 2 (this characteristic increases a lot with aging). Cross-culturally, it was indeed found that the ability to complete everyday tasks and to learn new information during the aging process were perceived to decrease with aging, in contrast to areas such as general knowledge and wisdom, which were perceived to increase. As general knowledge is expected to increase with aging, how does this relate to a decrease in the ability to learn new things? This question illustrates how different notions of aging (e.g. more traditional views related to wisdom and respect) and more modern views (e.g. related to cognitive decline) seem to contradict each other and hinder a lifelong learning perspective.
Focusing specifically on late-life language learning, Ramírez Gómez (2016) found that many foreign language teachers perpetrate these negative stereotypes in their teaching, as they assume that elderly people only take language courses for enjoyment, rather than expecting them to achieve certain goals (Andrew, 2012). In addition, stereotypes concerning cognitive decline also influence both teachers and older learners’ views vis-a-vis their abilities to learn during old age. As a result of these societal views, older learners may well feel less capable to learn a foreign language themselves (Andrew, 2012).
7 Within other fields of education, it has already been found that third-age learners’ perception of old age can be improved by taking part in healthy aging programs tailored towards seniors, such as University Programs for Older Adults (PUMAs) (Fernández-Ballesteros et al., 2013) or healthy aging workshops for community-dwelling older adults (Mendoza-Núñez, Sarmiento-Salmorán, Marín-Cortés, Martínez-Maldonado & Ruiz-Ramos, 2018). More generally, the study by Fernández-Ballesteros et al. (2013) indicates how their program departs from a lifelong learning perspective; hence their aim is to focus both on learning as well as altering people’s perceptions of aging and improving their well-being. The healthy aging workshops, on the other hand, were aimed at the possible prevention and control of chronic illnesses in older adults by improving perceptions of aging. The program was adapted to older learners more concretely, by making use of a textbook which was specifically designed for the project and comprised topics related to the biological, psychological and social aspects of aging (Mendoza-Núñez et al., 2018)
It follows that, in order to overcome ageist stereotypes with regard to learning, such paradigms or methods would need to be implemented in the foreign language classroom as well. In general, communication towards society and older adults should be geared towards debunking the myths related to learning in older adulthood (Ng et al., 2015). Therefore, what is currently needed is to replace counterproductive terminology with age-optimistic language in general and perhaps in the third-age language classroom in particular. As mentioned before, lifelong bilingualism has been attributed to several cognitive advantages (Alladi et al., 2013). In addition, several studies have investigated the process of learning a language in older adulthood, aimed at diminishing the general pattern of cognitive decline (Pfenninger & Polz, 2018; Ware et al., 2017). However, little attention is paid towards supporting narratives of lifelong learning. In a study by Matsumoto (2019), semi-structured interviews were conducted with seven Japanese (4 F, 3 M) third-age learners of Chinese, English, French, Italian and/or
8 Russian as an L2. The participants were between 56-76 years old. As the learners did not need to learn the second language for communication (having no opportunity to use it in their Japanese settings), career advancement or for living abroad, these initiatives exemplify the need for learning a language solely for personal enjoyment and well-being. In this sense, the study by Matsumoto (2019) looks at the process of third-age language learning from a Positive Psychology perspective (Macintyre, Gregersen, & Mercer, 2019). In the second part of this thesis, we will discuss how this perspective could help overcome ageist biases with regard to learning a language in older adulthood. As language learning is often seen as one of the most difficult new skills to learn (Bialystok & DePape, 2009), and with lifelong bilingualism known to promote cognitive aging, third-age language learning would be a particularly fruitful domain to investigate in relation to positive psychology and older adulthood.
This thesis thus aims to bring together two strands of research, by proposing an integration of perceptions of aging in the third-age language classroom. Building on previous bilingualism research, which has mostly focused on the prevention of cognitive decline, this thesis aims to put forward the multitude of beneficial effects of learning a language during the third age, among which socio-affective and linguistic effects. Contrary to earlier work, however, it aims to provide a broader societal perspective by looking into the effects of old-age related stereotypes and how to overcome these using Third-Age Language Learning (TALL) interventions. In this sense, this paper functions as an exploratory agenda for future research avenues within the field of TALL. As this is an emerging field, with more and more studies examining effects of language learning on old age, it would seem that at this stage it is beneficial to reflect on the foundations and perspectives of the field, starting with myths and misconceptions reflected in methodology and terminology. In the first part of this thesis, after having outlined the research questions and hypotheses for the thesis as a whole, the concept of the third age will be discussed from multiple angles, and a working definition will be
9 established. Terminology used to refer to the third-age life stage will be prominent in this first part. Following this, the origin and development of the University of the Third Age will be presented, exemplified by a few recent studies in this area. After this, the effects of TALL will be put forward as they have emerged from the scantly available studies to date. Firstly, the socio-affective effects of learning a foreign language during the third age will be discussed. Secondly, cognitive effects of TALL will be examined in order to indicate the cognitive benefits of learning a language during third age. Different types of intervention studies will be discussed and compared with foreign language learning interventions. In addition, theories of cognitive decline will be contrasted with non-decline theories. Thirdly, the linguistic effects of TALL will be addressed more specifically. As a final step in Part 1 of this thesis, we will look into perceptions of aging, outlining the effects of ageist stereotypes and the ambiguities within the healthy aging discourse. This is done in order to shed light on the role of preconceptions within third-age learning programs, and TALL programs in particular. Perceptions of aging then form the starting point of the second part of this thesis. In that second part, a research agenda will be put forward, detailing the possible applications for integrating perceptions of aging in the TALL Classroom. Firstly, the terminology of TALL will be discussed, as a consistent terminology to address concepts of later life will need to be established for this emerging field of study. Following this, the stereotypes embedded in TALL methodologies will be addressed and suggestions for future research will be given. Lastly, Positive Psychology (PP) will be introduced as a metatheory for future TALL research. The main aims of TALL will then be addressed from a PP perspective in order to provide a template for future research, proposing language learning in older adulthood as 1) a preventative tool for cognitive decline; 2) a means to overcome old-age loneliness and depression; and 3) an opportunity to stimulate lifelong learning. These topics are important for future investigations into the linguistic, social, and
10 cognitive effects of Third-Age Language Learning interventions, a field which is strongly emerging.
The Current Study
The current study thus consists of two parts: a state-of-the-art literature review on the topic of third-age learning and Third-Age Language Learning in particular, and a second part setting out a research agenda for years to come, most particularly pertaining to how positive terminology is pivotal in moving the field forward. Both parts are governed by a set of research questions.
Research Questions
1. Part 1: State of the Art of a Research Field
a. What are the socio-affective effects of TALL? b. What are the cognitive effects of TALL? c. What are the linguistic effects of TALL?
d. What are the stereotypes associated with learning in the third age?
2. Part 2: Research Agenda and Proposal
a. What terminology should be used within the field of TALL? b. What are the stereotypes embedded in TALL methodologies? c. How can Positive Psychology be applied to TALL research?
11 Hypotheses
Recent literature on third-age language learning has demonstrated that ageist stereotypes are hard to combat (Ramírez Gómez, 2016; Andrew, 2012). However, research within the field of healthy aging has illustrated that there are ways to improve (self-) perceptions of old age (Fernández-Ballesteros et al., 2013; Mendoza-Núñez, et al., 2018), although this line of work has never been directly linked to third-age language learning initiatives yet. We should thus find a way to tailor foreign language courses towards the specific strengths and needs of seniors and use positive terminology and tools to achieve this.
This could, for instance, be done by means of a thematic analysis of personal narratives of older adults taking part in a TALL program. In this way, common patterns could be derived by means of a coding process, so a profile of an effective foreign language teacher could be established. As put forward by Grotek (2017) this could help to capture the relevance of each language skill for the third-age language learner and seniors’ expectations of the teacher within the third-age foreign language classroom. Such a personal narrative method can be viewed as a suitable way to study older adults, as it “provides the subjects with an opportunity to freely share their individual views based on personal experiences and function as wise experts by being put in the position of giving advice to usually younger professionals” (Grotek, 2017, p. 132), which both align with positive stereotypes of aging (Settersten & Godlewski, 2016). “…and therefore may be appreciated by the elderly who are invited to participate in such research. Not surprisingly, the subjects of the present study explicitly report finding personal narratives more individualistic and less rigid than questionnaires that they had often been asked to fill in by researchers” (Grotek, 2017, p. 132). In addition, it can be helpful to identify specific language learning strategies and needs for seniors in order to be able to create appropriate instructional materials for teachers of third-age language learners (Pawlak, Derenowski & Mystkowska-Wiertelak, 2017).
12 Part 1: State of the Art
Defining the Third Age
In order to be able to answer our research questions, it is first of all necessary to look more closely at the term third age. ‘Third age’ is related to the period of time pertaining to healthy elderly individuals who are retired but still have a sense of “energy, excitement, purpose and well-being” (Oxford, 2017, p. 4). In order to comprehensively grasp the concept of ‘third age’, preceding stages of life also need to be defined. According to Moen (2011), the first age indicates the period of time during which a person is enrolled in education; the second age, then, pertains to the period during in which a person is employed. The third age is defined as the period of retirement and is most strictly defined by good physical health, as the fourth age would be the stage in which serious health problems start to occur. Where some demarcate the third age with a set number of years, others define it using a specified age of onset and termination, with different definitions varying between the ages of 50 up until 84 (Oxford, 2017). Both definitions could, however, be criticized, as aging is a very individual process, which is hardly generalizable and cannot be captured in numbers applied across the entire older adult lifespan. Whereas some individuals can reap the benefits of third age until late into their 90s, others may not be as fortunate (Andrew, 2012).
Furthermore, in order to develop a better understanding of the third age, it can also be very insightful to look at factors related to the aging process over the lifespan. In a longitudinal study by (Carstensen et al., 2011), it was found that emotional well-being improves over the life course. In addition, emotional experience was found to become more stabilized over time. Following Socio-emotional Selectivity Theory (SST) (Carstensen, 1995), the aging process is assumed to lead to emotional improvements. As elderly people become more aware of the time constraints to their life, they “become increasingly motivated to pursue emotionally meaningful goals and thus invest psychological and social resources to optimize emotional well-being”
13 (Carstensen et al., 2011, p. 12). Under this view, the third age is perhaps best defined from a more affective rather than a solely clinical perspective; the third age would then be the period during which people prioritize emotional well-being over other life goals.
Although Carstensen et al. (2011) thus suggest that elderly people experience more positive emotions during old age, this may not necessarily be the case for all people. Retirees need to adapt to the life changes that accompany retirement, as both physical and cognitive decline are likely to arise during this period of time, with perceived loss of societal usefulness (Oxford, 2017). Some of the main mental health issues for seniors are depression, anxiety and behavioural and mood disorders following dementias; 6 to 8% of the world population older than 65 has dementia, whereas this is around 25% for people aged 85 years and older (Knight, Kaskie, Shurgot & Dave, 2006). In addition, major depressive disorders seem to pertain to 1 to 2% of the elderly world population, but in clinical settings prevalence rates can rise to 40% in long-term care (Gatz, 2000). In the so-called Epidemiologic Catchment Area (ECA) study, 18,571 people were interviewed in order to study the prevalence rates of mental disorders and sociodemographic characteristics at five US National Institutes of Mental Health (NIMH). Based on these data (Regier et al., 1993), it was found that 5.5% of people over the age of 65 are diagnosed with an anxiety disorder. Yet, prevalence rates for both anxiety and depression were found to be significantly lower for people aged 65 and older in comparison to younger adults. It should, however, be taken into account that anxiety disorders can sometimes be underestimated in older adults (Lenze et al., 2000). Moreover, elderly have found to be less apt than younger adults at recognizing their symptoms for both anxiety and depression (Wetherell et al., 2009), which may point to a large proportion of elderly remaining undetected with regard to these mental health conditions.
14 The University of the Third Age
Due to the aging population, the idea of lifelong learning has received more and more interest, perhaps directly related to meaningful goals pertaining to old-age well-being. As pointed out before, research has indicated that learning can still take place during advanced life stages (Ramírez Gomez, 2016; Ramscar et al., 2014). In order to instigate the lifelong learning paradigm, the University of the Third Age (U3A) was established in 1972 in France (Formosa, 2014). Due to the enthusiasm following the initial lecture series, third-age education was continued in a more permanent way. Overarching aims were established, raising the importance of issues such as improving quality of life and creating a permanent program for third-age learners. At the start, U3As mostly followed a top-down approach. This meant that the university was in charge of the curriculum and academics had to be put on a pedestal for their intellectual reputation. However, when U3As were implemented in Britain in the 1980s, the educational method and underpinnings of third-age courses underwent considerable changes. In contrast to French third-age education, British U3As seemed to prefer a more bottom-up approach, further distancing academic courses for seniors from the university curriculum, but with more peer-to-peer contact. Over the last 30 years, a number of other models were established across the globe, ranging from ‘culturally-hybrid’ models, comprising both French and British components such as in Finland (Yenerall, 2003), to ‘French speaking North American’ types in Canada, where the U3A is part of the university, but the focus is on uniting higher education and third-age learning, rather than treating them as separate entities (Lemieux, 1995). Nowadays, more and more online U3A courses have been successfully developed and implemented in addition to physically-taught courses as well (Formosa, 2014).
An example of such contemporary U3A programs are the University Programs for Older Adults (Programas Universitarios para Mayores, PUMAs). PUMAs have now been set up all across the globe. In order to create coherence between the different PUMAs, they have also
15 been referred to as ‘black-box’ programs. This means that the courses at each university should be regarded as equal and university-independent as long as they are equal in duration across institutes. In general, the PUMA curriculum mostly consists of courses related to the humanities, history, literature, and health, with the exception of some courses in science and technology (Fernández-Ballesteros et al., 2013).
It is interesting to examine the effects that ensue from third-age courses. A recent study done into the effects of third-age education on active aging is the longitudinal PUMAs study that was conducted at the Autonomous University of Madrid from 2007 until 2010 (Fernández-Ballesteros, Molina, Schettini & del Rey, 2012). The sample consisted of a quasi-experimental group of 82 older adults and a passive quasi-control group of 76 people, matched for age, with both groups between the ages of 55 and 70. The quasi-experimental design can be attributed to the lack of randomly selected participants within the experiment. The quasi-experimental group was selected on the basis of exam results and the quasi-control group was selected from the Longitudinal Study of Active Aging (ELEA), initiated in 2006. The ELEA study examined the active aging of 458 inhabitants of Madrid aged between 55 and 75. About half (53.3%) of the participants in this study were retired, 10% was still working and 33.3% were housewives/homemakers (Fernández Ballesteros, 2011). The quasi-control group in the current study by Fernández-Ballesteros et al. (2012) had not taken part in any type of education during the research period but they were pre- and post-tested at the same points in time as the quasi-experimental group. The subjects in the quasi-quasi-experimental group attended a wide variety of courses, ranging from Spanish to Astronomy. The program took three years, comprising a total of 450 hours of lectures at the university for the quasi-experimental group. The Digit Symbol Substitution Test (DSST) taken from the Wechsler Adult Intelligence Scale (WAIS) (Wechsler, 1981) was used to assess memory and learning (cognitive functioning) outcomes for the quasi-experimental group. For this test, participants receive a type of legenda portraying symbols
16 which correspond to a certain number. The participant is then supposed to substitute as many numbers as possible by their corresponding symbol within a given time window. In addition to this cognitive assessment, affective measures were assessed by means of the PANAS scales, which is a self-report questionnaire consisting of two 10-item scales (rated on a 5-point scale) for positive and negative affect (Watson, Clark & Tellegen, 1988). With the quasi-experimental and quasi-control groups having been recruited from different studies, all measures were taken from the ELEA protocol in order to ensure that the same factors were studied for both groups. It was found that the quasi-experimental group increased their performance with regard to learning and memory, in contrast to the quasi-control group, whose participants showed a decline in their cognitive performance. In addition, participants enrolled in the PUMAs program experienced significantly more positive emotions in comparison to the start of the program, whereas the quasi-control group only less substantially decreased their negative feelings (Fernández-Ballesteros et al., 2012). As cognitive functioning has generally been found to decline with age (Gabryś-Barker, 2017), it is very significant to find that this trend could possibly be reversed by taking part in U3A programs. Also very important is the finding that U3A programs can positively modulate emotional well-being in seniors, as this has been related to increased cognitive functioning in older adult populations as well (Pot, Keijzer & de Bot, 2018). A different explanation for these results could, however, be a potential selection bias. As the people taking part in the PUMA program appear to be more highly educated than the quasi-control group (See Table 1 in Fernández-Ballesteros et al., 2012) and may have been more motivated to take part in such programs in general, they could potentially have improved more than the quasi-control group without having taken part in the program at all.
In a subsequent study by Fernández-Ballesteros et al. (2013), four PUMA initiatives in Chile, Cuba, Mexico, and Spain were studied over the course of one year. The researchers expected to observe improvements in emotional well-being, in line with the earlier findings.
17 The experimental group at the start comprised 313 people divided over all four countries. These participants were aged between 50 and 87 years old (M = 63.88) and took part in the PUMA program for one full academic year. The control group consisted of 190 people aged 50 to 91 years old (M = 65.24), who did not participate in any university program during the same period of time. In accordance with Fernández-Ballesteros et al. (2012), emotional well-being was measured by means of the PANAS scales (Watson, Clark & Tellegen, 1988), which was completed by both groups at the start and the end of the PUMA program. In line with expectations, it was found that well-being improved over the course of a year in the experimental group. By contrast, the control group did not show significant changes in well-being from the pre- to the post-test measurements. The authors do point to the great variation in PUMA courses that participants in the experimental group were enrolled in and which were not separately investigated. Although in theory these programs are regarded as equal due to their equal duration, in reality, the PUMAs take place at different institutes and do not provide the same course content. In addition, outcomes were averaged across the different PUMA institutes, so we do not know whether the outcomes were significant for each setting individually.
18 Literature Review (Effects) of Third-Age Language Learning (TALL)
Although Carstensen et al. (2011) claim that well-being generally appears to increase over the life course and during the third age specifically, the previous studies (Fernández-Ballesteros et al. 2012; 2013) have clearly indicated significant differences in well-being between participants who took part in the U3A programs and their control group counterparts who did not. This suggests that attending university courses in older adulthood could lead to improvements in well-being for seniors. However, as these programs were not specifically aimed at language, one might wonder how these results translate to language learning during the third age in particular. Indeed, there have been a number of studies that have hypothesized a very profound impact of third-age language courses in particular as opposed to other content courses (Bak, Long, Vega-Mendoza & Sorace, 2016; Pot, Porkert & Keijzer, 2019; Nijmeijer, van Tol, Aleman & Keijzer, forthc.). This is corroborated by the fact that, whereas the University of the Third Age has generally grown in popularity throughout the years, foreign language courses appear to be exceptionally popular in contrast to other types of courses (Gabryś-Barker, 2017). Foreign language courses in particular are so attractive to the third-age population because they provide a form of personal growth through what has been labelled ‘productive engagement’. As Park et al. (2014) illustrate in their study, productive engagement, which occurs when one is learning a new and challenging skill, should be separated from receptive engagement, which occurs when one is only socially involved in a new activity without learning a new skill. Indeed, Park et al. (2014) found cognitive improvements in memory, visuo-spatial processing, and processing speed for some of the productively engaged groups included in their study that were not attested in the receptive engagement groups. Although Park et al.’s study looked into the cognitive effects of digital photography and quilting in older adulthood, their results point at the possible effects of learning other types of skills, such as learning a language during the third age.
19 Third-Age Language Learning (TALL) is thus an emerging research field within the domain of adult L2 education. Therefore, according to Pot et al. (2019), only a few studies have looked into the socio-affective effects of TALL so far or indeed the cognitive effects to ensue from such courses. We will firstly go into the socio-affective effects of TALL in order to compare these outcomes with the effects of the PUMA courses discussed above. Then we will elaborate on the cognitive and linguistic effects of TALL in the ensuing sections.
The Socio-affective Effects of TALL
One study which looked into the socio-affective effects of TALL is a study by Pfenninger & Polz (2018). It is important to note here that this study looked at effects of language learning in subjects who had already learned a new language at an earlier stage in their lives versus those who had not. The experimental group consisted of six sequential German-Slovenian bilinguals aged 74-89 years old. The participants in the active control group were six Austrian monolinguals with minimal knowledge of any other languages than German, aged 63-84 years old. As the active ‘control’ group did not receive a different treatment or intervention than the experimental group; the only difference between the groups was based on their language background and prior language learning experience. No passive control group was included in the study. Participants in both groups had not learned any English before taking part in the course. The course consisted of one month of intensive English training, comprising a total of 24 hours of lessons. Socio-affective effects were studied by means of a questionnaire. The questionnaire looked at seven overarching outcome constructs: 1) the motivation to take part and 2) persist in the course; 3) self-perceived well-being; 4) the effect of the teaching method on perceived independence, communication abilities and self-efficacy 5) learning environment; 6) importance of the teacher; and 7) the type of language course. In addition, the Multidimensional Mood State Questionnaire (MDBF) (Steyer, Schwenkmezger, Notz, & Eid,
20 1997) was administered prior to and after each lesson. This instrument comprises 24 items that participants respond to on a 5 point Likert-scale; a high score points towards a positive mental state and a low score is indicative of a negative mental state. The MDBF did not show a significant increase in subjective well-being over the course of one month across either group, but overall well-being ratings were generally high throughout the entire course and before each class, which would not leave much room for substantial improvements. The socio-affective questionnaire revealed that 83% of the monolingual group reported a positive effect of the course on their social life. In addition, the bilingual group felt, too, that the English course provided them with the formation of social bonds and a fun group experience (no percentage was included for this group).
Another study looking into the socio-affective effects of TALL was conducted by Ware et al. (2017). The experimental group consisted of 14 French participants (M: age 75.42; SD: 8.93), with varying levels of English proficiency, taking part in a 4-month English course for older adults. Each lesson consisted of a scene from a television series from the 1950s and 60s. During the lesson the participants would receive a transcription of the scene and had to read a sentence out loud one by one and try to translate it. After the scene was translated, each participant had to read the sentence out loud once more. Then the scene was shown again, but this time with French subtitles. As part of this process, the movie was paused, so each line could be repeated - out loud - by the whole group. At the end, the participants could ask questions about what they had just learned. In order to measure loneliness and isolation, a French version of the University of California Loneliness Assessment (UCLA) was administered before and at the end of the course (De Grâce, Joshi, & Pelletier, 1993). The scores did not reveal a significant improvement in terms of loneliness and isolation over the course of 4 months, but in order to shed more light on this, semi-structured interviews were carried out individually with all of the participants after the course. From a subsequent analysis of these interviews, five themes were
21 distilled: 1) associations with school; 2) attitudes toward English; 3) motivation for learning English; 4) attitudes toward the program’s organization; and 5) social ties. One important finding was that: “All of the participants mentioned an absence of social bonding with fellow participants, and some regretted not being able to form stronger social ties with one another” (Ware et al., 2017, p. 7), which could possibly be a significant factor in the lack of improvement on the UCLA. However, this was not quantitatively measured within the study. The authors point to a few indicators for the lack of social bonding as reported during the interviews, such as variation in English proficiency levels and a potential generational difficulty with creating social bonds. It does need to be pointed out that these explanations are not necessarily deterministic, as one participant reported that the varying proficiency levels could have also supported interaction because of the need to ask each other for help (Ware et al., 2017). The length of the course could have played an additional role in the lack of significant outcomes, as the duration may have been too limited to develop social ties. However, this would not align with the outcomes of Pfenninger & Polz (2018) with only one month of training. Moreover, the content and set-up of the course itself could have played a significant role as well, as it did not seem to allow for a lot of communication during the lesson. Social engagement has to be an inextricable part of a foreign language course, if one of the intended learning outcomes is indeed the ability to communicate in that language. However, the TALL classroom would have to provide an explicit focus on social activities in order for meaningful communication to actually take place, perhaps especially in the case of elderly people. This was mostly lacking in the study’s set-up.
A recently published systematic review by Klimova & Pikhart (2020) looked into seven studies related to the social well-being but also the cognitive effects of TALL, among which the previously discussed studies by Ware et al. (2017) and Pfenninger & Polz (2018). It was found that, as TALL courses provide possibilities to meet other people, this could lead to
22 improvements in older adults’ well-being. In addition, they claim that “It is partly through the stimulation of social well-being that the cognitive effects of FLL might be observed” (Klimova & Pikhart, 2020, p. 7).
In short, the previous studies seem to suggest that the formation of social ties plays an important role in the subjective well-being that older learners experience during a foreign language course. However, more research should be done in order to investigate how social bonding and subjective well-being are linked within the third-age language learning classroom and how this can and should be an intricate part of any third-age language learning curriculum. Moreover, we should not exclude the possibility that well-being effects could need a longer time frame to become noticeable beyond the scope of a TALL course. Therefore, it would be important to also include follow-up measurements a number of months after the course has ended.
The Cognitive Effects of TALL
Both physical and cognitive decline are documented for older adulthood. With regard to physical problems, we may encounter a decline in sight and hearing abilities and sensory acuity in a more general sense; with regard to cognitive decline, what is most frequently reported is memory loss (Oxford, 2017) and processing speed reductions (Wingfield & Grossman, 2006). TALL, in particular, is a relevant domain to study in this respect, as bilingualism has shown to be advantageous with regard to different cognitive functions, such as monitoring, shifting, and/or inhibition (Bialystok, Craik, Klein, & Viswanathan, 2004), which is perhaps most pronounced at a life stage where these facets may be compromised (Bialystok, 2016). A relevant question to ask, then, is whether effects similar to those induced by lifelong bilingualism can be reached as part of foreign language courses instantiated in older
23 adulthood. In the next section, we will firstly provide more detail about the first study looking into cognitive effects associated with bilingualism by Bialystok et al. (2004) in order to contextualize current research into the cognitive effects of TALL.
In the now seminal paper by Bialystok et al. (2004), a study was done into the differences in inhibitory control in monolingual vs bilingual participants. A Simon task was performed by different age groups. The Simon task is a conflict monitoring task said to tap especially into inhibition as a cognitive mechanism. During the task, stimuli of a specific colour are shown at the left or right side of a computer screen. Each colour is linked to a key on the corresponding side of the keyboard. The participant is then supposed to press this key when the colour appears on the computer screen (congruent). However, sometimes, the stimulus position changes to the opposing side of the screen (incongruent). Incongruent items generally lead to longer response latencies (RTs). The ‘Simon effect’ should be interpreted as the average difference in response times between the congruent and the incongruent trials. 40 participants took part in the study, the first half of whom were between 30 and 54 years of age (M = 43). The other half were seniors aged 60-88 years old (M = 71.9). Each group consisted of ten monolingual English speakers from Canada and ten Tamil-English bilinguals from India. It was found that, on average, the bilinguals outperformed the monolinguals with regard to RTs and showed a smaller Simon effect. Strikingly, bilinguals from the older age group also appeared to make fewer mistakes than their monolingual counterparts. Although the Simon effect appeared to be smaller for the younger adults in general, bilingual older adults still showed a smaller age-related increase in their Simon effect than their monolingual age-matched controls. In the second study reported in that same paper, a larger sample of participants was included, again consisting of two groups: 64 middle-aged participants aged 30-58 years old (M = 42.6) and 30 older adult participants between 60 and 80 years of age (M = 70.3). This time, the stimuli that formed part of the Simon task increased from two to four, thus increasing the task demands.
24 Again, bilinguals outperformed the monolinguals with regard to RTs and showed a smaller Simon effect across the two groups. The bilingual older adults again showed a smaller Simon effect than their monolingual peers. In the third study, which was essentially a training study, results from the Simon task indicated that the monolinguals in this study ultimately performed on a par with their bilingual peers, given enough practice, whereas the bilinguals continuously performed better throughout the complete task. It should be noted, however, that the bilinguals in the study were simultaneous bilinguals, who had consistently used several languages from the age of ten or younger. It follows that these results do not necessarily extend to sequential bilinguals or second language learners in older adulthood. In a plethora of research on cognitive effects to emerge from lifelong bilingualism, a robust finding is that lifelong bilingualism may delay the onset of dementia with 4.5 years on average, which is of particular relevance to the third-age population (Alladi et al., 2013; for a recent systematic review see Brini et al., 2020). The question arises whether such cognitive effects may be extended to language learning in older adulthood as well.
In the studies by Ware et al. (2017) and Pfenninger & Polz (2018), apart from the socio-affective effects of third-age language learning, cognitive performance was studied as well. In Ware et al. (2017), the included cognitive measure comprised a French adaptation of the Montreal Cognitive Assessment (MoCA) (Nasreddine et al., 2005). No significant improvements were observed with regard to cognitive functioning over a four-month period of English training. However, it should be noted that the MoCA scores of the participants were already generally high prior to the course (M = 26.46 out of 30), which did not provide much room for improvement as a function of the language course. Hence, the insignificant improvements could also be interpreted as an indication of cognitive maintenance, which still surpasses a scenario of cognitive decline. In spite of this, the level of English proficiency at the start of the program differed significantly across participants, ranging from beginner to
25 advanced proficiency levels of English, which could have largely impacted the cognitive outcomes of the training (see Bak et al., 2016). As the study was mostly qualitative in nature, no control group(s) were included in the study, which also limits the generalizability of the outcomes, corroborated by the small sample size of 14 participants.
Pfenninger & Polz (2018) examined late-life language learning effects in a group of subjects who were already bilingual versus a group who were functionally monolingual and had never systematically studied a foreign language at an earlier stage in their lives. In the study, a Stroop task (Stroop, 1935) was used in order to measure inhibitory control; and the concentration test for geriatric patients (AKT) (Gatterer, Fischer, Simanyi, & Danielczyk, 1989) was used to measure attention and concentration. In addition to these functions, the AKT has also been used to provide an indication for the stage of dementia, due to changes in the ability to concentrate. Over the course of a month of language training, significant effects were found for the Stroop task for both groups, but the monolingual group outperformed the bilingual group with regard to inhibitory control, both prior to and after the course had ended. In addition, no significant effects could be observed with regard to attention and concentration for the bilingual group, in contrast to the monolingual group, who significantly improved over the course of the training. These findings illustrate that bilinguals do not always outperform monolinguals when it comes to inhibitory performance. Furthermore, with regard to attention, it could have been expected that the bilingual group would not improve as much as the monolingual group, because they are expected to cognitively function at a higher level prior to the English training due to their bilingualism, but this was not the case. In order to improve the reliability and interpretation of these findings, the study emphasizes the importance of including both a passive and an active control group in future studies. This suggestion is underscored by Park et al. (2014), who had already illustrated the beneficial effects of a productively engaging activity
26 over a receptively engaging activity with regard to improvements in memory function in old age.
A study incorporating both an active and passive control group in investigating the effects of language learning in seniors is Nijmeijer et al. (forthc.). This protocol paper details a study design set to investigate the cognitive effects of different types of interventions related to musical learning (active control) and social interaction by means of creative workshops (passive control) alongside a foreign language intervention as primary cognitive intervention. The study examines a group of seniors with memory complaints based on self-perceived measures of decline. Musical learning is included in this study as the active control intervention as musical learning and processing are said to rely on similar neural networks as language (Jäncke, 2009). However, the key difference between the two interventions, and what is used as the main reason to label language learning as the primary experimental intervention tool, is the fact that learning a second language leads to interference with the first language and therefore cognitive flexibility is required to resolve interference effects between these languages (cf. Bice & Kroll, 2015). It follows that cognitive interventions without these supposed competition effects do not train cognitive flexibility to the same extent as a language intervention does. Thus, the language intervention is expected to lead to the largest improvements in cognitive flexibility. As both music and language interventions have been found to improve well-being (Pfenninger & Polz, 2018; Ware et al., 2017; Seinfeld, Figueroa, Ortiz-Gil, & Sanchez-Vives, 2013), which could in turn impact cognitive performance (Pot et al., 2018), a social intervention is included to discern the effects of the different interventions. The differences between the groups in relation to well-being are hypothesized to be insignificant, as previous studies have illustrated how the different types of interventions all led to improvements with regard to well-being due to their social component of group interaction. Nevertheless, the research design can be regarded as
27 very innovative within the field of TALL as it has taken preventative measures to overcome the limitations that emerged in earlier studies (see Pfenninger & Polz, 2018).
In addition to these studies, a study by Bak et al. (2016) studied the interrelationships between language proficiency levels and cognitive abilities, but not specifically related to seniors. It particularly focused on the interactions of Gaelic language proficiency levels and attentional functions. The study consisted of 67 participants aged 18-78, with one experimental group (n = 33); one active control group (n = 16) and one passive control group (n = 18). The participants in the experimental group took part in a one-week intense Scottish-Gaelic course and were divided into three proficiency groups: complete beginners (n = 15), elementary (n = 8) and intermediate (n = 10) learners. The courses were tailored towards the learners’ proficiency level. In addition, an active control group was included who took part in courses of the same length and intensity but with a different content (either CELTA (English teaching) training, art courses, or a documentary film course), as well as a passive control group who continued with their normal daily routines without any intervention but were tested at the same time as the intervention groups. The cognitive measures that were used were part of the Test of Everyday Attention (TEA) (Robertson, Ward, Ridgeway, & Nimmo-Smith, 1994), which was administered in order to measure attentional functions such as inhibition and switching. In the experimental group, performance on attentional switching improved across all proficiency levels and age groups. However, the beginner group showed a larger effect than the intermediate group. The authors ascribe this difference to the research by Park et al. (2014), illustrating that “novel, cognitively challenging activities seem to be more effective in improving performance on tests of cognitive functions, such as working memory, than less taxing, familiar ones” (Bak et al., 2016, p. 8). In addition, the attested improvements were long-lasting, as a follow-up test nine months later indicated sustained improvements for participants who had continued to study
28 Gaelic for five or more hours per week. In contrast, the control groups did not improve significantly on attentional switching.
Contrary to the studies on TALL by Pfenninger & Polz (2018) and Ware et al. (2017), the study by Bak et al. (2016) takes into account the effects of the different language proficiency levels of the participants in their findings. As novelty appears to play an important role in cognitive improvements, it is pivotal to take into account the baseline proficiency levels in the target language, as this could indicate to what extent each proficiency group is expected to improve cognitively as a result of third-age language training programs. On the other hand, socio-affective effects do not have to be dependent on language proficiency levels. It is therefore as pivotal to take into consideration the individual goals of third-age language learners prior to their participation in the language course, as older learners are also much more varied in their reasons to learn a second language than other age groups (see Matsumoto, 2019).
Although the previous studies mostly depart from a cognitive decline point of view and assess how a language intervention may combat this age-related decline, different studies have tried to reject this idea by providing non-decline perspectives. As Oxford (2017, p. 6) argues: “Though cognitive information processing often takes longer for third agers, this problem might be balanced by existing verbal knowledge and positive biographical, lifestyle and contextual factors (e.g. greater education, consistent physical activity and age-supportive social practices)”. In other words, although cognitive decline and lower processing speed might influence our learning process during third age, this does not mean that learning cannot take place anymore (Ramírez Gómez, 2016; Ramscar et al., 2014). Ramscar et al. (2014) even go as far as to claim that healthy seniors do not show cognitive decline when it comes to information processing: according to Ramscar et al. (2014), the phenomenon of adults having to take longer in their search for words can be explained by the fact that, as vocabulary grows with experience over the course of a lifetime, so too does the time needed to sort through our memory. This, in
29 turn, leads to increased difficulty and reduced speed with information processing. Rather than labelling this process as cognitive decline, Ramscar et al. claim, it is more accurately described as the result of accumulated knowledge. In addition, research has pointed out that some aspects of memory can still be actively engaged during older adulthood. An example in point is memory strategies. Whereas elderly frequently experience and report negative conceptions in relation to their memory capacities, they may actually have developed very helpful strategies to overcome these problems on a higher level by means of ‘metamemory’. This concept refers to the knowledge and regulation of memory processes. Metamemory, in turn, is a part of ‘metacognition’, which is the knowledge of the human mind and cognition in general (Sigelman & Rider, 2014).
In a study by Hines, Touron & Hertzog (2009), it was found that older adults can still monitor their memory in order to support their learning of semantically unrelated word pairs. In the study, a group of younger adults aged 18 to 25 (M = 21.8) was compared with a group of older adults aged 60 to 75 (M = 66.8). Prior to the study, the participants filled out a questionnaire, performed the Shipley Institute of Living Scale Revised Vocabulary Test (Zachary & Shipley, 1986), and the Digit Symbol Substitution Test (DSST), taken from the Wechsler Adult Intelligence Scale (WAIS) (Wechsler, 1981). The experiment consisted of two study-test stages in which the participants performed an associative recognition task. As the authors explain: “The experiment was a 2 (Age: Young, Old) × 2 (Response Time Feedback: Given, Withheld) × 2 (Study-Test Phase: 1, 2) mixed design, with age and feedback as between-subjects independent variables and study-test phase as the within-between-subjects variable” (Hines et al., 2009, p. 6). In the first phase, the experimenter had allocated a certain study time to a given word-pair; in the second phase, participants self-paced their study time for each item. In each stage, 60 semantically unrelated word pairs were presented on the screen. After this, the participants gave a Judgement of Learning (JOL) in order to judge their ability to memorize the
30 items (on a continuous scale from 0 to 100% confidence) ten minutes after completion of the study round. As a next step, the participants would test themselves in a self-paced manner. In order to test their memory, they were instructed to press a key to indicate whether or not they had learned the word-pair before. One half of each age group was provided with Reaction Time (RT) feedback, so their RT in seconds would be presented on the screen after each response, whereas the other half was not. The latter group would instead receive a blank screen for the same duration as the screen in the feedback condition. In the feedback group, feedback was only provided for the odd-numbered trials. After the even-numbered trials, participants were to guess their RT from a list of nine categories. This would then be followed by a Confidence Judgement (CJ) in order to judge their confidence level in relation to the correctness of their previous estimation. It was found that JOLs and CJs influenced the self-allocated study time of the participants in phase 2: when participants were less confident about their judgement in phase 1, they would spend more time on that item in phase 2. These findings thus indicate a relationship between study time allocation and metacognitive processes that was similar for both age groups. Moreover, “Older adults showed a robust effect of JOLs and CJs on phase 2 study time, and CJs actually had a greater influence on study time for older adults relative to younger adults” (Hines et al., 2009, p. 13).
In conclusion, the current literature on the cognitive effects which ensue from TALL programs are inconclusive. As studies have used different cognitive measures, they have also investigated different types of effects. We could thus both expect cognitive maintenance effects as well as improvements depending on the different factors which were studied, also taking into account methodological challenges of earlier work (e.g. small sample size and relatively high cognitive abilities prior to the start of the intervention). Furthermore, it is important to contrast perspectives of cognitive decline with non-decline viewpoints. From this latter perspective, we could be more mindful of the cognitive strategies seniors have already developed throughout
31 their life, such as verbal knowledge and metamemory, in order to not only see the weaknesses but also cater towards the strengths of seniors within the TALL classroom.
The Linguistic Effects of TALL
In addition to the socio-affective and cognitive effects of third-age language learning, it is equally relevant to examine the linguistic effects of learning a language during older adulthood. Departing from a non-decline perspective, learning a language can be a very meaningful and educational experience for seniors, due to the nature of language itself. Whereas the cognitive aspects of learning a second language may be of particular interest to many researchers, it is unlikely to be the top priority for seniors themselves who enrol in a foreign language course. Indeed, older adults may want to use language as a tool to communicate with others. Moreover, linguistic gains made throughout the course could also be used as a means to combat ageist stereotypes related to learning in old age, as it would illustrate that even after a certain age a language goal can still be achieved by taking part in a language program. By extension, the (indirect) influence of linguistic improvements on socio-affective and cognitive domains should also not be overlooked.
A pilot study by Kliesch, Giroud, Pfenninger & Meyer (2018) studied the English language proficiency gains of ten participants aged 65-74. The participants were native speakers of German and had no or only basic knowledge of the English language prior to the course. The course consisted of 60 English lessons, of which at least one lesson per day was a self-study lesson completed at home. The administered language tests to measure language gains were a C-test to assess general L2 proficiency, the Hueber assessment test Next A1 in order to measure vocabulary gains, as well as reading and listening skills, and an oral translation test. The three tests were administered prior to the course (T1) and directly after the course had ended (T2).
32 Overall, the study found participants to improve their general English language proficiency over the course of a three-week training. However, as the scores were averaged across the three tests, no conclusions can be drawn about improvements for specific language domains or progress of individual learners across these domains.
In a study by Pfenninger & Polz (2018) following the pilot by Kliesch et al. (2018), linguistic improvements were found in areas such as (linguistic) self-confidence and communicative skills over the course of a month of intensive training. Whereas the term ‘linguistic self-confidence’ is not specifically defined by the authors, it seems to be related to the confidence that participants gained from improving their English skills by taking part in the course and their overall willingness to communicate (cf. Piechurska-Kuciel, 2014). The study included 12 participants aged 63-90 without any prior knowledge of English. As the training in the pilot study by Kliesch et al. (2018) had previously been found to be too demanding for the age group, a four-week course of 12 lessons in total (instead of a three-week course consisting of 60 lessons) was devised. The language tests which were administered were the C-test to assess “grammatical production skills” (Pfenninger & Polz, 2018, p. 4) and two tasks to study receptive vocabulary. These were administered halfway during the course (T2) and after the course had ended (T3). In addition, a Foreign Language Learning questionnaire was administered prior to and after each lesson. At the beginning of the course, a language background questionnaire was administered. The study does not present the FL proficiency results for the pre-test (at T1) (see table 1 of the study). However, the authors do seem to indicate to have tested participants at T1, at least for the C-test. Although a language background questionnaire was administered prior to the start of the course, it would also be important to have an objective baseline for English language proficiency, as the participants’ receptive or implicit knowledge could be larger than they would explicitly be able to report. In contrast to the study by Kliesch et al. (2018), non-significant improvements were found for the
33 receptive vocabulary tasks and the C-test between T2 and T3. Plausibly, the different moments of testing and smaller number of lessons included in the training could have influenced these results.
In contrast to the study by Pfenninger & Polz (2018), two of the “monolingual German native speakers” (Kliesch et al., 2018, p. 54) in Kliesch’ study appeared to be reasonably proficient in at least one other language than German, which raises the question what the authors actually denote when using the terms ‘monolingual’ and ‘bilingual’(see also Surrain & Luk, 2017 for more details about which terms to use with regard to bilingualism). As the study did not comprise a control group, the internal validity of the study could also be questioned. As a general point of attention, the previously mentioned studies into the linguistic effects of TALL comprise sample sizes that are quite small, which can be problematic for their generalizability towards the larger third-age population.
One study which used a much larger sample (n = 105), is a study by Pikhart & Klimova (2020), examining the well-being and linguistic effects of TALL. A questionnaire using a six-point Likert scale was administered consisting of twenty statements regarding participants’ self-perception of their improvements regarding their L2 proficiency. In addition, tests were carried out to objectively measure participants’ vocabulary and grammar gains (although they were deemed irrelevant for the study’s objectives). Indeed, it was found that the seniors did not improve much with regard to their vocabulary size and grammatical proficiency. However, even though the participants accurately perceived their improvements in the L2, these results contrasted starkly with people’s enjoyment of the course.
Nevertheless, variations in the tests used to assess linguistic improvement over time between the different studies can negatively impact the generalizability of research findings on this topic. Table 1 below details this variability as well as outcomes of the scantly available studies in this domain. As corroborated by a review of the methodological concerns of the field
34 by Van der Ploeg, Keijzer, & Lowie (2020), improvements in language proficiency are only rarely taken into account within TALL research and should be more widely implemented in future research.
Table 1
Linguistic improvements due to Third-Age Language Learning
Linguistic improvements
Study Groups (n per
group), (gender) Experimental group: Age in Years Control group(s): Age in Years (linguistic) self-confidence and communicative skills
Pfenninger & Polz (2018) 1 experimental (6) and 1 active control (6), (8F, 4M) 74-88 63-84 General L2 proficiency Kliesch et al. (2018) 1 experimental (10), (6M, 4F) 65-74 - Grammar and vocabulary Pikhart & Klimova (2020) 1 experimental (105), 1 young control (102), 1 elderly control (102), (exp.: 68F, 37M) 55-80+ Young control: 19-23, elderly control: 55-80+
