Reframing the Cultivation Metaphor in Education by Reflecting on CLIL Professional Development Through a Forest Gardening Lens

1 Introduction

In education, cultivation metaphors are deeply ingrained in the conceptualization of teaching and learning processes (Mintz, 2018). Teachers are often likened to gardeners, charged with the responsibility of nurturing young minds to bloom and flourish. Such metaphors underscore the role of educators as individuals who meticulously tend to their students, planting seeds of knowledge and providing the necessary care and support for growth. In a similar vein, professional development (PD) facilitators are often seen as cultivators of teacher expertise, guiding them through workshops and training sessions to refine their skills. However, these conventional gardening metaphors perpetuate notions of top-down practices and standardization (Zipory, 2020) while overlooking the complexity and interconnectedness inherent in educational landscapes.

To describe and understand phenomena as intricate and nuanced as PD, more holistic and integrated ways of viewing teachers and their learning are needed. To this end, various systems approaches put forth by researchers highlight the complex relationships and interdependencies present in education. Unlike traditional linear models, ecological systems approaches and complex dynamical systems approaches (Ehrenfeld, 2022; Garner & Kaplan, 2021; Koopmans, 2020; Larsen-Freeman, 2023), for example, acknowledge the reciprocal relationships and feedback loops that shape teaching and learning environments, including PD.

To continue this shift toward a more integrated, systems model of teacher PD, it is essential to reconceptualize the cultivation metaphor still prevalent in education. The forest garden metaphor presents a compelling alternative to the traditional cultivation metaphor by emphasizing intentionality, localization, agency, diversity, growth from decay, and especially, interdependence. In a forest garden, a variety of specifically chosen plants coexist, each contributing to the health and productivity of the ecosystem. Similarly, teacher PD should embrace diversity in approaches, perspectives, and methodologies, recognizing that different students, teachers, and schools have unique needs and strengths. By fostering a supportive and collaborative environment, akin to the interconnectedness of a forest garden ecosystem, a thriving community can be cultivated where there is continual learning, reflection, and growth.

In subsequent sections, deeper exploration will be conducted into the principles of forest gardening and how those principles can inform the design and implementation of PD and related research in education. Although this is primarily a conceptual paper, I draw upon personal experiences and empirical data from a cross-national content and language integrated learning (CLIL) PD experience called Project SCILLA. The application of the forest garden metaphor in this manner serves as a lens to assess Project SCILLA beyond the traditional PD measures of teacher classroom practices, teacher beliefs, and student achievement (Guskey, 2000). Additionally, this application provides a glimpse into the usefulness of the forest garden as a metaphorical lens for PD as a component within a dynamic and sustainable education system.

2 Overview of a Cross-National PD Program

Since President Nazarbayev’s call for trilingualism in 2007, Kazakhstan has been experimenting with its education policies, specifically those related to medium of instruction. In an effort to develop English proficiency alongside Russian and Kazakh languages, teachers at the secondary and post-secondary level are expected to use English to teach various subjects, including biology, chemistry, physics, and computer science (Karabassova, 2022). The transition to teaching technical content using English as a medium of instruction has been met with numerous challenges (Karabassova, 2020, 2021, 2022; Manan et al., 2024), hence the call for PD and teacher education program development throughout Kazakhstan.

To this end, the team leaders of what would be called Project SCILLA drafted a proposal in 2019 for an American Councils grant-funded PD program. The initial program was to include an in-person workshop phase in Kazakhstan during the summer of 2020. As the COVID-19 pandemic worsened in early 2020, it was clear that the program would need to shift to a virtual format. The revised program maintained the objectives of increasing programmatic, methodological, and technological capacity for CLIL pedagogies in science, technology, engineering, and mathematics (STEM) courses in rural Kazakhstani universities. The phases of the program shifted to include a series of virtual workshops throughout the summer of 2020 for the 30 participants from 11 universities. The program culminated with participants designing, implementing, and reflecting on a CLIL-informed online module during the fall of 2020. For many of the participants, this module was a single lesson, but for others, it was a series of lessons or the revamping of a course syllabus.

Throughout the six-month program, participants were supported individually and in small groups by the U.S. and Kazakhstani team leaders. To continually adapt the curriculum and PD structure, ongoing feedback was collected in the form of exit tickets, questionnaires, focus groups, and interviews. Data sources drawn upon in this paper include recordings from workshop sessions, focus groups, and interviews as well as artifacts, including lesson plans and other participant-produced documents.

3 The Magic and Misuse of Metaphor

The “SCILLA” in Project SCILLA’s name stands for “STEM Content Integrated Language-Learning Activities” and was deliberately wordsmithed to represent scilla plants. Scilla is a genus of plant species, some of which are native to Central Asia (Iannotti, 2021). In this way, SCILLA symbolizes the scientific realm which is connected to our work with STEM teachers, and it symbolizes sustained growth. Scilla are hardy perennials that do not need to be re-planted each season; they are easily naturalized, meaning that they self-propagate and can thrive in places where they are not native. Our team thought this was an apt metaphor to describe the cross-national, multilingual PD that we set out to do: we would be the gardeners planting scilla (i. e., CLIL pedagogies) that would, ideally, continue to grow and spread after our PD concluded.

Thinking about teachers, teacher educators, or in our case, PD facilitators, as gardeners is hardly unique. As Lakoff and Johnson (1980) argue, such metaphors are more than literary devices comparing two things that are not usually juxtaposed: metaphors are a way of life. We make sense of the world’s complexities through metaphorical conceptual systems. Accordingly, education researchers have used metaphor as a means of exploring teachers’ understanding of teaching and learning (e.g., Oxford et al., 1998; Saban et al., 2007; Worden-Chambers, 2020). After all, “there is no way to talk about education (or almost everything else, for that matter) without using metaphors” (Zipory, 2020, p. 45).

Considering that the German word Kindergarten translates to “garden of children,” it comes as little surprise that Western education has been influenced by horticultural metaphors. Mintz (2018) details how the gardener-educator metaphor has been invoked by Western philosophers of education, including Jean-Jacques Rousseau, Noah Webster, John Locke, and Horace Mann. Gardening and cultivation metaphors have been used to describe both teacher-centered pedagogies in which a teacher (gardener) has the power and authority to mold their students (plants) and learner-centered pedagogies in which a teacher fosters the individual growth of their students by creating optimal environments. It is worth noting, as scholars point out, that these metaphors prevail in non-Western contexts as well (e.g., Saban et al., 2007; Tao, 2017; Yurtseven, 2017).

To understand processes as complex as teaching and learning, gardening metaphors have been leveraged not only by philosophers but also by teachers. Oxford et al. (1998) analyzed teacher narratives and found that “teacher as nurturer” was a common metaphor associated with a “learner-centered growth perspective” (p. 40). Two Turkish studies also revealed the existence of such metaphors among pre-service and in-service teachers. In their analysis of pre-service teachers’ responses to the prompt, “A teacher is like ... because ...,” Saban et al. (2007) reported that 9% chose metaphors that represented “teacher as nurturer/cultivator” (p. 131). Likewise, the most frequently used metaphors to describe “professional development” among in-service Turkish teachers were those categorized as “a growing/evolving entity.” Within that category, teachers frequently used the metaphors of “tree” and “seedling” to reflect their perceptions of PD (Yurtseven, 2017).

As magical as it may seem to envision teachers as hard-working gardeners who foster the growth of society’s most fragile, Zipory (2020) points out the dangers of using cultivation metaphors in education. He argues that “cultivation metaphors exclude, or even eliminate with almost no trace, alternative modes of uncontrolled or ‘wild’ learning” akin to how agriculture and sedentism have marginalized other ways of living (p. 45). He further argues that by using cultivation metaphors, characteristics of agriculture’s grain economy, such as “repetitiveness, coercion, rational management, and lack of diversity” are imposed onto education (p. 53).

Although metaphors cannot exactly match what they are describing, they have the capacity to (re)frame our thinking and our actions (Lakoff and Johnson 1980). Thus, it is necessary to re-consider the metaphors we use so that they are not more “deluding than illuminating” (Cook-Sather, 2003, p. 950). If we see metaphors as a way to bridge the divide between theory and practice in education, we should find the most suitable ones. Zipory (2020) concludes his critique of the cultivation metaphor by suggesting we turn to the forest. He asks that we consider the “irony, wonder, and the unpredictability of self-governing ecosystems represented by the forest” (p. 59). Rather than focus on the linearity, standardization, and commodification associated with traditional gardening and agriculture, the forest could serve as a more interconnected, interdependent, and harmonious ecosystem mindset.

To combine Zipory’s (2020) suggestion of the “wild” forest with the reality of some of the “unwild” practices in education, I propose the metaphor of the forest garden to better understand education systems, and more specifically, the potential of PD within those systems. Rather than directly theorize the role, design, and implementation of PD, I hope to harness the symbolic power of metaphor to evoke vivid imagery and build a new model to conceptualize PD within complex multilingual, multicultural settings.

4 Overview of the Forest Garden

According to Breyer (2020), “rewilding” and “ungardening” refer to the return of manicured gardens to more natural, forested environments with native species that attract native pollinators and other wildlife. As ecological and in line with Zipory’s (2020) call as this sounds, these terms did not seem like a good fit for the work of Project SCILLA and other PD programs that adopt “non-native” curriculum. Project SCILLA was rooted in CLIL pedagogies and the teaching of STEM content through English, which has become synonymous with trilingualism in Kazakhstan (Karabassova, 2020). Knowing the power and hegemony that the English language exudes and its “non-native” status in Kazakhstan, a return to the “wild” or the “native” did not make metaphorical sense.

Thus, when I came across the following definition of the forest garden from horticulturist Robert Hart (1991/1996), I was attracted to the liminality of the concept—it is neither a natural forest nor a traditional garden:

Like the natural forest, it comprises a wide diversity of plants, occupying seven levels or “stories,” but, unlike that natural forest, almost all its plants have been carefully chosen to meet human needs. It is, in fact, an attempt to create a model life-support system, which would enable a family or small community to achieve a considerable degree of self-sufficiency ... while enjoying health-giving exercise in a beautiful, unpolluted, and stimulating environment. (p. xiii)

Forest gardening shares practices with the land use system of agroforestry as well as principles from permaculture. Bukowski and Munsell (2018) explain that “both disciplines are adaptable, and adaptability is critical in a community food forest [forest garden] because the demands of intercropped trees and plants change over time across unique site conditions” (p. 117). In addition to being dynamic, forest gardens foster agency since their human creators bear the responsibility of intentionally designing systems that meet community needs.

The seven layers of the forest garden include the canopy, the low tree layer, the shrub layer, the herbaceous layer, the rhizosphere, the soil surface layer, and the vertical layer. Another distinguishing feature of the forest garden is the mindful use of available technologies, such as trellises, solar ponds, irrigation devices, and terracing techniques (Hart, 1991/1996). Ideally, these layers and technologies work together to form a self-perpetuating, self-fertilizing, self-watering, self-mulching and weed suppressing, self-pollinating, self-healing system that is resistant to pests and diseases. In short, forest gardens are healthy ecosystems that provide for and respond to those who nurture and sustain them.

While the metaphor of the forest garden has its own shortcomings, I believe it can be a launching point for conceptualizing the intricate ecosystems that exist in education globally. In the following sections, I explore the key principles of forest gardening and parallels in education systems by highlighting the role of teacher PD. To this end, I refer to a model of dynamic and sustainable professional development (henceforth, DSPD) which incorporates principles and concepts from forest gardening (Figure 1).

Figure 1

Dynamic and Sustainable Professional Development (DSPD) Model

5 PD as Part of an Interconnected Ecosystem

As Cook-Sather (2003) warns, the more strictly we impose a metaphor onto education, the less useful it may become. Thus, rather than delineating each level of the forest garden and its equivalent in education systems (e.g., global rhetoric is like the canopy layer, national policy is like the low tree layer, teachers are like gardeners, and so forth), the DSPD model keeps the principles of forest gardening at its flower-shaped core. Each petal of the flower represents a key principle of forest gardening that should be practiced when designing, implementing, and evaluating PD. The flower is situated within two circular figures to acknowledge the influence of stakeholders and the education policy landscape. The DSPD model aims to recognize the multiple levels at play, but rather than focus on the hierarchy that is often present across and between those levels, this model idealistically highlights interconnectedness.

The two outer circles (policy landscape and stakeholders) of the DSPD model give a nod to the interconnectedness that is inherent in most systems models. Nested frameworks, such as Bronfenbrenner’s (1979) ecological systems theory and Ricento and Hornberger’s (1996) language planning and policy (LPP) onion metaphor, inevitably create hierarchy with outer layers being larger than, and encompassing, inner layers. However, Ricento and Hornberger note that such “multilayered” constructs have “agents, levels, and processes [that] permeate and interact with each other in multiple and complex ways” (p. 419). To represent such permeation and interaction, the policy figures in the DSPD model have arrow-shaped heads leading into one another, and the stakeholder figures are represented with dashed lines.

The arrow-shaped heads of “Global Rhetoric,” “National Policy,” and “Local Policy” in the DSPD model insinuate the potential cyclical nature of policy. In contrast to the LPP onion model, the DSPD model does not nest the various levels of policy; instead, global rhetoric, national policy, and local policy occupy the same outer layer. While this collapsing of levels does not account for all the nuances of education policy landscapes, it is idealistic in envisioning a landscape where policy at any level has the potential to influence policy at another level due to the agency and advocacy of its stakeholders.

For this reason, the stakeholder figures in the DSPD model are represented with dashed lines rather than being bounded entities. The categories of “Teachers,” “Students,” “Administrators,” “Politicians,” “Professional Development Facilitators,” “Parents,” and “Other Stakeholders” are distinct in some ways but often overlap and intersect, especially through the use of agency. For instance, a teacher may leverage their agency by advocating for or against particular policies. While they may not take on the official role of elected politician, through advocacy, voting, and/or their classroom practices, teachers have political influence. Therefore, the dashed lines in the DSPD model allow space for the blurring of boundaries that occurs in interconnected systems.

To highlight these aspects of the DSPD model, the Kazakhstani language-in-education policy landscape provides fertile ground for investigation. In Kazakhstan, global rhetoric touting English as a lingua franca is interconnected with national language-in-education policy and, subsequently, local policies aimed at implementing national mandates. More specifically, a trilingual policy (Kazakh, Russian, and English) was introduced in 2007. While CLIL was not a nationally-mandated pedagogical approach, it was adopted in various contexts, including in a government-supported pilot policy project at the Nazarbayev Intellectual Schools (NIS) as well as at other mainstream schools.

In her comparative study of this language-in-education policy implementation at NIS and mainstream schools, Karabassova (2021) found that teachers in each context had differing experiences. Even though the top-down policy was nearly identical in each context, the PD provided and the speed of implementation varied. The teachers at the well-funded NIS were given more time, training, and autonomy, whereas the teachers at mainstream schools had fewer resources yet were expected to implement CLIL approaches almost overnight with minimal training.

In short, contextual realities result in uneven implementation of national policies, and these local differences can lead to subsequent policy change. Since teachers are typically the local actors who negotiate education policies in the classroom on a daily basis (García & Menken, 2010), the category of “local policy” should include the practices that occur behind closed classroom doors. Consequently, teachers can use their agency to implement policies with or without fidelity. Karabassova (2021) found, for example, that whether or not teachers had sufficient English proficiency and CLIL training, they were required to teach in English if they wanted to remain in the profession.

With some teachers continuing to teach in Kazakh instead of English (Karabassova, 2022), it may come as little surprise that the national language-in-education policy was reformed in 2019. President Nazarbayev urged the Ministry of Education and Science not to rush trilingualism with “clumsy” methods and a lack of appropriate teacher training (Kazakhstan Today, 2019). Such national policy shifts, likely influenced by local policy realities, in turn affect global rhetoric. Many people are keeping a close eye on the situation in Kazakhstan as they continue researching and speculating on the feasibility of trilingual policies (e.g., Kambatyrova, 2024; Li, 2022) as well as the efficacy of CLIL pedagogies in various contexts (e.g., Banegas & Hemmi, 2021; Van Mensel et al., 2020). In this manner, global rhetoric, national policy, and local policy are not only interconnected but also interdependent.

6 Stakeholders’ Fluid and Independent Roles

Since teachers, students, administrators, PD facilitators, and other stakeholders work within (and sometimes push back against) the confines of the education policy landscape, the DSPD model nests them inside the outer policy layer. Just as the canopy and tree layers of the forest garden shade the plants below, the interconnected policy landscape also strongly determines the kinds of practices that can grow within its shadows. This is not to say that stakeholders are passive players; on the contrary, they hold the power to shape their experiences and shift the policy landscape through their actions.

As previously mentioned, the dashed lines around each stakeholder in the DSPD model intentionally highlight the fluidity of stakeholders’ roles. We often talk about teachers as lifelong learners, but so too are students, administrators, and policymakers. Likewise, students, administrators, and policymakers can take on teaching roles. For instance, in the two contexts that Karabassova (2022) investigated (the elite NIS and mainstream schools), teachers expressed a desire to improve their own language skills and often relied on their students. Karabassova further explained that such “partnerships” between teachers and students not only resulted in students correcting their teacher or translating peers’ answers but also led to a friendlier classroom atmosphere in which students had increased power (p. 1540). In this way, the roles of “teacher” and “student” became less distinguishable.

Reflecting on Project SCILLA, we too experienced a blurring of boundaries between PD stakeholders. One of our Kazakhstani team leaders served as both a PD facilitator and as a teacher PD participant. She was a core team member who was able to process PD content for her own circumstances while also considering where her peers may need additional scaffolding according to their backgrounds and contexts. The insight she drew upon from her unique positionality was crucial for a cross-national PD program like ours. Her localized knowledge substantially influenced the PD program; for example, we reduced the number of tasks and changed the types of assignments based on her perception of the other participants’ workloads and workplaces, which were ever-evolving during the COVID-19 pandemic.

For these reasons, it does not make sense to think about stakeholders as bounded islands within an ecosystem; nor does it make sense to think about some stakeholders as gardeners who are fostering the growth of other stakeholders. The interdependencies and fluid roles of PD stakeholders create more reciprocal and nuanced relationships. While there are discernible power hierarchies between these various stakeholders, the DSPD model makes space for imagining fewer boundaries between these roles. Harrison’s (1992) characterization of the forest as “a place where the logic of distinction goes astray” (p. x) and Zipory’s (2020) call for the “distinctions between student and teacher” to be “questioned and transformed” (p. 59) should push us to continue (re)envisioning the boundaries and relationships between stakeholders in education systems.

7 Localization and Agency Through Intentional Planning and Design

At first glance, a forest garden might not have the appearance of a garden at all. Without linearity and uniformity, its complexities are not apparent to those who do not understand its design. As Hart (1991/1996) elucidates, the forest garden is “often a precise, multistoried structure put together by people who had an intimate knowledge of the properties, products, and habits of growth of the plants involved” (p. 7). Like forest gardens, components of education systems, including policies, curricula, assessments, and PD, should be carefully selected and planned for in consideration of the local context and the system as a whole. Additionally, those who hold intimate knowledge of the local context should be involved throughout the selection, planning, and implementation processes.

As Watson (1996) explains, a forest garden “requires thoughtful planning at its inception, and lots of work to get it planted and well established” (p. 69). Part of this planning involves the selection and placement of plants. These selections should be made by considering not only the characteristics of the various plant species, but also characteristics of the local environment as well as the needs of those who live in that environment. The DSPD model was designed with a sun in the middle (between the petals of the flower) to remind us of the importance of local conditions. In the same way that plants should be selected based on localized conditions, like the amount of sunlight in a particular area, so too should components of education systems be selected.

The intentional selection of components may not make sense to those outside the system. Just as ornamental, or non-native, plants can be part of a healthy forest garden ecosystem, “non-native” or “borrowed” policies, pedagogies, and curricula can sometimes be part of an intentional, localized education system. Whether or not we agree that English language and CLIL pedagogies have a place in Kazakhstan, their presence is not accidental. Global rhetoric touting the prestige of English and its value as a lingua franca is sure to have influenced Nazarbayev’s desire for a trilingual nation.

Although some may argue that English is “invasive” since it threatens the vitality of local languages, its pervasiveness signals shifting environments. Non-native plants can thrive in new areas because climatic conditions have changed due to phenomena like industrialization, urbanization, and global warming (Bitner, 2017). Likewise, advances in transportation, telecommunication, and technology shape societies in ways that may encourage and support multilingualism as well as a need for a lingua franca, like English. Of course, further advances with artificial intelligence and yet-to-be-created technologies will impact dynamic ecosystems in undetermined ways.

No matter what components are intentionally selected for a system, stakeholders from all groups should be involved throughout the planning, implementation, and evaluation processes. When provided opportunities to share one’s expertise, more appropriate selections can be made. Additionally, stakeholders can feel a sense of agency. Rather than become “passive adopters” (Karabassova, 2020, p. 46) of top-down policies, stakeholders can feel empowered to engage in policy implementation. As we know, it is teachers who enact policies in their classrooms ( García & Menken, 2010). Like gardeners who best know the lay of the land, teachers have an intimate understanding of their students and the local context that most policymakers do not. Thus, providing space for teachers’ autonomy and agency may increase their motivation and lead to more policy “buy-in” (Karabassova, 2021; Manan et al., 2024).

In the case of Project SCILLA, several of the U.S. team members had localized knowledge from visiting and/or living in Kazakhstan. More importantly, the Kazakhstani team members were “on the ground” and able to provide information and advice to the entire team about the local contexts; this was crucial with the continual shifts caused by the pandemic. In this way, our local team members were able to anticipate challenges and localize the PD structure and content to address our participants’ changing needs and desires.

Throughout the PD, we also attempted to make space for participants’ personal and collective agency. Since we worked with teachers and administrators from 11 universities, we had to alter our content and expectations to recognize their contextual realities. To this end, we utilized exit tickets, questionnaires, focus groups, and interviews to collect feedback, which was used to adapt the PD workshop content, assignments, and the overall project timeline. During the small group sessions, we also made space for participants to share their experiences and advice with each other in a dialogic format.

Just as it is important to choose plants that will thrive in a particular environment, localized and meaningful PD must be intentionally incorporated into education systems. A less top-down approach that involves active participation from all stakeholders allows for knowledge sharing and permeation of ideas across groups (Lo, 2020; Ricento & Hornberger, 1996). Such agentive collaboration can, ideally, lead to more appropriate and effective policy, curricula, pedagogy, and practices.

8 Creating Dynamic and Self-Sustainable Systems Via (Re)Evaluation

While there may be much human intervention at the fore, a forest garden is designed to become “a largely self-regulating system” (Watson, 1996, p. 69). An important part of creating an ecosystem with diverse, compatible plants is ongoing research. Hart (1991/1996) says that as “a forest garden ... changes from year to year, even from day to day,” “I would urge anyone who starts a forest garden to adopt a creative attitude towards it; to learn and observe, to study and do research” (p. 150). In a similar manner, education stakeholders at all levels should continually reflect on their role(s) within their ecosystem, and PD should be seen as one element of an ongoing process of teacher professional growth and subsequent educational change.

Neither plant growth nor teacher change is linear. Some plants, like Swiss chard or asparagus, can be cut and produce multiple harvests within one season; perennial plants return each season; and the decomposition of dead plants is a crucial part of the carbon cycle, which, ultimately, leads to more growth. Teacher professional growth is also a complex process that is often cyclical (Clarke & Hollingsworth, 2002; Imants & Van der Wal, 2020; Lo, 2020). Thus, as teachers and other stakeholders grow within a system that is also evolving, (re)evaluation becomes necessary.

When comparing the processes of trilingual education implementation in two different contexts, Karabassova (2021) found that a gradual transition was more conducive to supporting teachers and creating capacity for CLIL pedagogies. The continuous process of rolling out CLIL pedagogies provided time for teachers to improve their own language proficiency and to reflect on the changes being made. Feedback from PD facilitators on video recordings of the teachers’ actual teaching was one specific strategy used to promote ongoing teacher growth in Karabassova’s study and in Project SCILLA. Ricento and Hornberger (1996) suggest that action research is another powerful way for teachers and students to evaluate policy at the classroom level.

Project SCILLA built in opportunities for facilitators and participants to reflect on their learning throughout the PD. The leadership team met weekly to strengthen relationships between team members, discuss research aspects of the PD, evaluate participants’ (and our own) progress, and plan for upcoming sessions. In addition to receiving feedback from the PD facilitators throughout their CLIL-informed online module development, teacher participants evaluated their own progress by updating individualized Gantt charts. The Gantt charts included space for participants to create a work timeline, detail their lesson planning process, connect their work to CLIL principles (Ball et al., 2016), and reflect on their successes and challenges throughout the PD. While some participants did not regularly update their Gantt charts, others continued to use them after the conclusion of the PD. One participant shared that she has introduced Gantt charts to her students as a self-reflective, organizational tool. In this small way, we can see an aspect of the PD that has been adopted and adapted to become part of one participant’s dynamic education ecosystem. Only time will tell if such deliberately cultivated tools and pedagogies will remain as resilient as well-established forest gardens.

9 The Beauty of Ordered Diversity and Creative Practices

For an intentional, localized forest garden to become dynamic and sustainable, diversity is the key. Hart (1991/1996) calls for “an ordered diversity, governed by the principles and laws of plant symbiosis; all plants must be compatible with each other” (p. xvi). It is not unusual for a forest garden to have dozens of species of plants within a relatively small space. Such diversity has several purposes: to provide food continuously through the year; to prevent pests or diseases from spoiling the entire garden; and to contribute to the aesthetic and spiritual aspects of the land.

In education systems, such diversity might be thought of as the array of policies, curricula, and pedagogies present rather than the demographic composition (e.g., age, sex, race, religion) of stakeholders in that system. More centralized education systems, like that of Kazakhstan (OECD, 2018), tend to have less plurality in terms of policy, pedagogy and curriculum than decentralized systems. While centralized systems have some advantages, Zipory (2020) warns of the “control and coercion” that is often present (p. 2). Just as “cultivation is aimed at diminishing diversity” with the goal of producing “identical, sturdy seed, whose grown crops are visible, long lasting, and easily commodified” (p. 54), standardization in education may lead to similar outcomes.

Rather than specializing in one instructional approach or one specific school of pedagogy, it may be wiser to diversify approaches available to administrators, PD facilitators, and teachers. Accordingly, there will be more opportunity for stakeholders to choose what is most appropriate for their context. In the case of Project SCILLA, we incorporated aspects of four frameworks that have been used for CLIL PD purposes (Lo, 2020): sheltered instruction observation protocol (SIOP); content-language-literacy integration for curriculum planning; the contextualization, awareness, practice, and autonomy (CAPA) sequence; and genre-based pedagogy and multimodalities-entextualization cycle (MEC). Depending on the audience (content teachers vs. language teachers) and context (elementary vs. secondary; rural vs. urban), one of these four frameworks, or a particular combination of them, was more appropriate for some participants than others.

While it may seem messy implementing multiple approaches within one context, the value of diversity cannot be overlooked. An education system and its stakeholders can become more dynamic, sustainable, intentional, and localized when plurality is valued. When given choices, there is an opportunity for creativity in the selection and adaptation of educational components. In this manner, choosing an appropriate policy, curriculum, or pedagogy is not just about how it “looks” but also how it “feels.”

In the ways that forest gardening merges science and art, so too, should education systems. Horticulturist Morrison argues that a landscape must be “experientially rich, beyond the visual dimension” (Roach, 2021, para. 17). He explains that ethereal qualities, such as luminosity and aroma, create landscapes that evoke our spirituality. Hart (1991/1996) agrees that “the forest garden is far more than a system for supplying mankind’s material needs. It is a way of life and it also supplies people’s spiritual needs by its beauty and the wealth of wildlife that it attracts” (p. xvi). How might education systems change if we viewed them as sites of creative and spiritual possibility rather than sites of “artificial incorporation” (Karabassova, 2022, p. 1543) of top-down policies?

10 The Synergy of Symbiosis

Hart (1991/1996) uses the term “ordered diversity” to emphasize the need for forest garden species to not only be compatible with each other but, ideally, form symbiotic relationships (p. xvi). He claims that “symbiosis—‘living together,’ or mutual aid—is the basic law of life” (p. 8). While some of these complex interactions between organisms are understood, “there are innumerable interactions like these that science has not yet discovered” (p. 13).

In education, some relationships may also be described as symbiotic. Take teachers and students, for example, whose relationships are defined by give-and-take interactions. In a traditional initiate-response-evaluate model of questioning, a teacher makes a point by building off a student’s response; likewise, a student receives feedback from the teacher on their response. This interaction can be mutually beneficial, or symbiotic. As previously mentioned, Karabassova (2022) found partnerships between teachers and students in CLIL classrooms that could be described as symbiotic; by positioning themselves as learners and receiving feedback from students, teachers were able to work on their language proficiency while students demonstrated their own.

Similarly, curricula can have symbiotic relationships wherein the teaching of one subject benefits another and vice versa. At its core, CLIL might be considered symbiotic since it is built upon “a dual-focused educational approach in which an additional language is used for the learning and teaching of both content and language” (Coyle et al., 2010, p. 1). Working in the area of second language teacher education, Banegas (2022) argues that synergies between two curricula (in the L1 and L2) can provide “learners with an educational experience that goes beyond specific subjects and languages” (p. 386). Such an experience is in line with Hart’s (1991/1996) “spiritual element [of evolution] which ensures that the whole is more than the sum of its parts” (p. 8).

In Project SCILLA, we observed symbiosis in various ways. For example, several participants transferred their learning from the PD to local secondary teachers. One participant shared, “I am thinking not about my [university] students, actually. I’m thinking about the school subject teachers ... This CLIL project is not for me, it is for them. I am preparing myself for them” (Interview, January 28, 2021). As an assigned mentor for local teachers, this participant saw Project SCILLA as an opportunity for her to grow as a PD facilitator and help those whom she described as being in more dire need of intervention. This organic process of content from our PD propagating in other areas we had not anticipated may be akin to symbiosis and the self-healing potential of the forest garden. As Hart (1991/1996) says, “companion plants ... affect their neighbors by stimulating their growth or warding off pests and diseases” (p. 11). Similarly, we saw our PD participants growing while simultaneously stimulating the growth of their colleagues.

11 Mindful Integration of Technology

A forest garden requires much initial planning, but how much work is required once it has been established? Hart (1991/1996) argues that “like the natural forest, it is a largely self-regulating, developing ecosystem that requires minimal maintenance” (pp. 50-51). Maintenance can be minimized with intentional planning and the creative and meaningful use of available technologies. In the forest garden, this could be as simple as using mulch to prevent weeds or as complex as an irrigation system.

In education, sustainable systems also require a certain amount of work and benefit from the creative and meaningful use of technologies and pedagogies. Although PD should be sustained, with some researchers advocating for programs of at least two years’ duration (Darling-Hammond et al., 2017), it might also be thought of as seasonal or interspersed work. Teachers are busy and often overworked, so PD must be intentionally scheduled and structured. Furthermore, the technologies and pedagogies introduced to teachers via PD must be compatible with their realities; potential benefits of incorporating PD content may be disregarded if teachers, and subsequently, their students, do not buy into it.

Once the Project SCILLA team realized that our PD would be virtual, we made conscious decisions about which digital tools to use, knowing that these would serve as models for our participants. Working closely with our Kazakhstani team members, we decided to use Zoom video conferencing for workshops as well as Slido (live polling and quizzes), Jamboard (interactive whiteboard), and collaborative Google Suite applications. We also used WhatsApp to communicate since this was a commonly used platform in Kazakhstan.

We found that many participants incorporated the digital tools and related pedagogies demonstrated by team leaders into their CLIL online modules. Slido, Jamboard, and breakout room features in Zoom were popular choices among our participants for facilitating student-centered CLIL activities that moved from private to public (Ball et al., 2016). Furthermore, some participants made creative adaptations to utilize technology in ways that were best for their local context (e. g., using BigBlueButton instead of Zoom for video conferencing; using WhatsApp for creating small groups instead of Zoom breakout rooms). Such adaptations highlight the intentionality and localization needed to integrate technology in a mindful manner.

As discussed more thoroughly in Novitskaya et al. (this issue), technology integration not only requires a deep knowledge of technology but also of content, pedagogy, and the ways in which they all interact (TPACK; Koehler & Mishra, 2009). Unfortunately, the realities of the COVID-19 pandemic forced our participants and their students to adopt digital tools more quickly than they may have been ready for. Consequently, the implementation of digital tools and associated pedagogies was difficult; technology-related issues were the most prevalent challenges reported by our participants when conducting their online modules. One participant shared, for example, that she did not have the necessary technological and pedagogical knowledge to choose appropriate online alternatives for familiar face-to-face practices, like running dictation. This illustrates the importance of teachers positioning themselves as lifelong learners as well as the pertinence of deliberate and thoughtful technology integration.

12 Thoughts on Weeding and Allowing for Growth From Decay

Before a forest gardener integrates technology to prevent the overgrowth of weeds, Hart (1991/1996) recommends reconsidering the place of weeds which can have “beneficial, symbiotic effects on neighboring plants, to which they impart something of their health, strength, and vigor” (p. 72). His “most basic rule of pruning is: ‘If in doubt, don’t.’ It is far better to underprune than overprune, especially in the case of plants in their first, tender years” (p. 77).

If we transfer this notion to education systems, we might reconsider the place of practices that do not perfectly align with policies or mandated curricula and pedagogies. Rather than eradicate deviant practices, we ought to reflect on the reason for their presence as well as their impact. In the case of the implementation of Kazakhstan’s trilingual education policy, Karabassova (2022) found instances of teachers “artificially incorporating English language into their teaching” (p. 1543). The reasons for this included low English proficiency (of teachers and students), low teacher self-efficacy, a lack of understanding of CLIL pedagogies, and a desire to meet formal policy requirements.

While some practices of “artificial incorporation” may be harmful to students, others may be beneficial. The practice of first teaching in the L1 and/or L2 (Kazakh/Russian) and then duplicating the entire lesson in the L3 (English) noted by Karabassova (2022) is more likely to be harmful since it is time-consuming and not focused on authentic language learning. On the other hand, if a teacher is translanguaging and using the L1 and/or L2 throughout a lesson that is supposed to solely use the L3 as the medium of instruction, there could be benefits for the teacher and students. Lin and He (2017) argue that “translanguaging can function as both pedagogical scaffolding strategies and opportunities to negotiate and affirm students’ identities and build teacher-student rapport” and thus should be “regarded as an essential pedagogical strategy in CLIL classrooms, especially with students of multicultural and multilingual backgrounds” (p. 237).

In Project SCILLA, we noted the use of L1 and/or L2 in many of our participants’ online CLIL modules. Although we discussed the potential benefits and appropriate place of translanguaging in CLIL pedagogies, there was still uncertainty and guilt surrounding the fluid use of languages in the CLIL classroom (see Green-Eneix & De Costa, this issue). Perhaps our leadership team should have done more to encourage “weeds,” such as translanguaging practices, to thrive. For example, Lin and He’s (2017) study of translanguaging in CLIL classrooms in Hong Kong could have served as a catalyst for discussion of similar practices in our participants’ local contexts.

As counterintuitive as “ungardening” may seem, Hart (1991/1996) reminds us that when he “continued to work—or rather not-work” his herb garden, he found that the plants “could look after themselves very nicely” (pp. 46-47). We, too, would be wise to reconsider the place of practices that initially appear to be “undesirable” and trust the instincts of those who know the local ecosystem best.

In the forest garden, the relationship between photosynthetic plants and fungi may be the epitome of symbiosis: dead plant material is decomposed by fungi which in turn provide nutrients to other plants. This example of growth from decay is an important lesson for education stakeholders. What we consider failure could turn out to be a powerful learning opportunity. Thus, the “recycling of every waste product” (Hart, 1991/1996, p. 12) should be a goal of policy analysis and curriculum evaluation.

In Kazakhstan, the rollout of trilingual education policy was not as successful as anticipated. Rather than start anew, a requirement of 15 universities to teach four STEM subjects in English was changed to “at least one out of the four [subjects]” (Karabassova, 2020, p. 45). Further adjustments were made when President Tokayev announced a focus on only Kazakh and Russian languages. Instead of considering this an outright policy failure, there is an opportunity to (re)integrate trilingual policy into spaces with more established bilingual policies and practices. Individual schools or regions in Kazakhstan that have successful Kazakh-Russian bilingual education models might be appropriate spaces to pilot trilingual policies and practices. In this way, there may be growth from decay which can lead to evolution of an even more “complex, integrated organism” (Hart, 1991/1996, p. 8).

13 Reaping the Harvests of Reframing Cultivation Metaphors in Education

With intentional planning and the creative use of technologies, like greenhouses, forest garden communities can reap the fruits of their labor year-round, even in temperate climates. It is easy to recognize such harvests as successful outcomes of forest gardens, yet there are so many other possible benefits. If we view forest gardening as a philosophy, we can see that “the way of life these forest gardens provide is secure, healthy, cooperative, constructive, and creative” (Hart, 1991/1996, p. 6). Thus, the value of the forest garden stretches far beyond dietary and monetary yields.

Likewise, when we think about the bounties of education systems, we should not merely consider standardized test scores and graduation rates as measures of success. When evaluating the effectiveness of PD, it is common to analyze teacher knowledge, teacher practices, and student achievement (Guskey, 2000). Knowing that teacher professional growth and subsequent student learning are not linear processes (Clarke & Hollingsworth, 2002; Lo, 2020), what might be overlooked when we limit ourselves to such measures? If we instead view PD as an interconnected and interrelated part of a dynamic ecosystem, we can begin to appreciate the covert outcomes brought about by intervention at one level of the system.

Although the Project SCILLA team gathered data regarding teacher knowledge, beliefs, and practices throughout the PD, that data revealed only a snapshot of participants’ cyclical growth. Our data captured participants’ experiences throughout the six-month PD and up to three months after teaching their CLIL online module. If the average rate of knowledge decay is just 37 days, even after intense PD (Liu & Phelps, 2020), continuing to follow up with our participants may not reveal sustained knowledge. For this reason, the residual outcomes that we had not planned for perhaps make an even more compelling case regarding the potential of PD interventions.

Just as scilla plants propagate, so too, does Project SCILLA live on, even if its detailed content may have been forgotten. As mentioned previously, several participants adapted materials and activities from our PD to use with local secondary teachers in a training-of-trainers model. Other participants noted that the PD has led to: consumption of CLIL resources that we had recommended; sharing and advocacy of PD content with department chairs; creation of a CLIL handbook for colleagues; and even the design of a new TESOL course at one of the participant’s universities. In these ways, participants may (or may not) forget knowledge from Project SCILLA, but they are replacing or supplementing it with new, related knowledge appropriate for their local ecosystem.

The impact of PD may transcend all layers of the education system though we may not observe many of the interactions and potential changes for years, if ever. Thus, the analysis of PD at the level of teachers (knowledge, beliefs and practices) and students (achievement) is akin to measuring the quantity and quality of a single, seasonal crop. Without stepping back and considering the entire ecosystem and its nuances over time, we miss valuable information. Rather than succumb to the “dark side” of cultivation metaphors in education which impose agricultural values of “repetitiveness, coercion, rational management, and lack of diversity” (Zipory, 2020, p. 53), we should look to the philosophies and practices of forest gardening which shed light on other possibilities in education systems.

With that said, there is no single forest garden model. Rather than attempt to create education systems that mirror the layers of a forest garden, we should emulate the forest garden’s principles of sustainability, interconnectedness, intentionality, localization, agency, diversity, creativity, symbiosis, mindful integration of technology, and growth from decay as outlined in the DSPD model (Figure 1). By analyzing Project SCILLA and other CLIL PD in and outside of Kazakhstan, we can see the potential of the forest garden metaphor to conceptualize the relationships and power dynamics between and across various stakeholders in education in response to global rhetoric, national and local policies, and interpersonal interactions. Furthermore, the forest garden metaphor provides a unique lens to examine a variety of contexts, particularly multilingual, multicultural, and/or cross-national ones, due to its emphasis on plurality, adaptability, and contextualization.