Two Routes to Depth: Deep Learning in Indonesian and Uzbek Educational Philosophy

Introduction

Global discourse on educational quality has increasingly focused on deep learning, an approach that moves beyond mechanical rote memorisation to cultivate meaningful understanding, critical thinking, and the ability to transfer knowledge across life contexts (McPhail, 2021; Kovač et al., 2025). The term carries an unavoidable ambiguity. In computer science it names a family of neural network architectures. In educational research it names a quality of learner engagement. This article uses it exclusively in the second sense, and adopts the Indonesian policy rendering pembelajaran mendalam as a clarifying cue.

Indonesia has positioned itself at the forefront of this conversation. Deep learning was formalised as national policy through Regulation of the Minister of Primary and Secondary Education No. 13 of 2025, which mandates the approach across schools from the 2025/2026 academic year (Permendikdasmen No. 13, 2025). The regulation forms part of an integrated set that includes the graduate competency standard and the content standard (Permendikdasmen No. 10, 2025; Permendikdasmen No. 12, 2025). The agenda has also become an instrument of educational diplomacy. On 7 November 2025, Indonesia’s Minister of Primary and Secondary Education, Abdul Mu’ti, presented the deep learning framework during a bilateral meeting with Uzbekistan’s Minister of Preschool and School Education, Ezozkhon Karimova, in Tashkent (Kemendikdasmen, 2025). Minister Mu’ti described deep learning as an educational paradigm aimed at conscious, meaningful, and joyful learning experiences, enabling students to reflect upon and contextualise knowledge across life situations (Kemendikdasmen, 2025).

The exchange occurred between two countries with markedly different educational landscapes. At that meeting the Indonesian minister described a system of almost 55 million primary and secondary students, 4.2 million teachers, and more than 400,000 schools, spread across some 17,000 islands and around 700 regional languages, while his Uzbek counterpart described roughly 14,000 schools serving about 9 million students with some 700,000 teachers and mentors (Kemendikdasmen, 2025). These are the figures the two ministers exchanged, and they are reported here as such; national statistical series compiled on different definitions give somewhat different totals, with the Indonesian education database recording 52.9 million students in 2024/2025 and the Indonesian language agency verifying 718 regional languages (Badan Pengembangan dan Pembinaan Bahasa, 2024; Kemendikdasmen, 2025b). Uzbekistan’s system is meanwhile undergoing active reform, with teacher working conditions and professional development recently documented at scale (OECD, 2024) and preschool provision substantially restructured over the past decade (UNICEF, 2017; Mamadaliyeva & Qosimova, 2024). Together the two countries present an instructive case for philosophical examination of the foundations of educational partnership.

Existing scholarship on deep learning divides into three broad positions. The first treats the concept as coherent but under-supported by curriculum architecture, and argues that depth follows from disciplinary coherence rather than from pedagogical enthusiasm (McPhail, 2021). The second is sceptical, holding that the concept is definitionally unstable and carries an unexamined cultural bias towards Anglophone progressivism (Kovač et al., 2025), a position that philosophical framing has extended into existential, ethical, and epistemological registers (Faizah et al., 2025). The third is empirical and concerned with conditions of implementation, documenting how teachers negotiate structure and agency as they move towards deep learning pedagogy (Lim et al., 2016) and how the approach is being received in Indonesian primary education (Feri et al., 2025). The first position identifies what depth requires but says little about culture. The second identifies the cultural problem but stops at critique. The third shows that institutional conditions mediate the concept without asking whether different institutional conditions can deliver it equally well. That last question motivates this article.

Scholarship on the two traditions compared here has developed along separate paths. In the Indonesian context, Dewantara’s educational philosophy, particularly the Sistem Among and the principle of kemerdekaan (freedom), has been analysed as an indigenous educational philosophy (Efendy, 2023; Basuni et al., 2021; Suwignyo, 2021; Budi Wijaya et al., 2024; Azhar Fuadi et al., 2025), read through postmodern epistemology (Ramadan, 2025), located within the wider field of philosophy of education (Siswadi & Murtiningsih, 2024), mobilised in critiques of contemporary legislation (Prabtama & Dwi Kartika, 2024), and positioned by Tilaar (2002) within the longer trajectory of Indonesian educational and cultural thought. Uzbekistan’s post-independence reforms have been examined through analyses of cultural and spiritual revitalisation (Sharipov, 2022; Akhmedova, 2024), the integration of classical thinkers (Uzakova, 2025), theoretical work on the philosophy of education produced within the region (Kundizalieva, 2023), the recovery of pre-Soviet Central Asian pedagogical figures such as Abdukadir Shakuri (Ashurova, 2025), and the structure of policy governance (Gafurov, 2024; Mavlonov & Bo’riyev, 2025). Studies on Dewantara rarely engage in cross-cultural philosophical dialogue beyond Southeast Asia, while analyses of Uzbek reform remain dominated by policy, historical, or sociological perspectives (Ozawa et al., 2024), with explicit philosophical examination still limited.

Research Gap

Three research gaps follow from this state of the literature. The first concerns the absence of systematic efforts to compare the philosophical foundations of Dewantara’s thought and Uzbekistan’s educational reforms within a single analytical framework. Scholarship on each has developed considerably, yet the two bodies of work do not speak to one another. The second concerns the disconnection between deep learning as contemporary policy and deep learning as a philosophical concept. This review located no study that critically links Indonesia’s deep learning agenda to its local philosophical roots through Dewantara, and none that compares such a linkage with the philosophical foundations of reform elsewhere. The third concerns the lack of a philosophical foundation for the newly formalised Indonesia-Uzbekistan educational cooperation. The bilateral dialogue is currently driven by exchange of policy practice rather than by a synthesis capable of sustaining a long-term partnership. These gaps become more pressing in the context of Hebert’s (2023) call for comparative educational philosophy to move beyond Western-centred paradigms through decolonial approaches.

This article addresses these gaps by exploring deep learning through a comparative philosophical analysis of Dewantara’s educational philosophy and the philosophical foundations of Uzbekistan’s educational reform. It identifies and reconstructs the core principles of both traditions, analyses their convergence and divergence in conceptualising the nature, purpose, and process of deep learning, and reflects on what this synthesis implies for bilateral cooperation. The central argument is that deep learning is not a culturally neutral concept. It requires grounding in local philosophical traditions to achieve authentic and sustainable implementation, and comparative dialogue between non-Western traditions can enrich global understanding of meaningful education.

The specific yield of the comparison deserves statement at the outset, since a paired reading must justify itself against two independent single-case studies. Read alone, Dewantara supplies a humanistic account of learner freedom that is easily assimilated into the progressivist canon. Read alone, Uzbek reform supplies a familiar story of post-Soviet state-led modernisation. Read together against a common yardstick, the two cases show that identical dimensions of deep learning can be reached through opposite institutional logics. One proceeds from the child outward to the community; the other proceeds from the state inward to the classroom. That equifinality is invisible in either case taken by itself. It is also what disciplines the claim that deep learning has no single culturally proper delivery mechanism.

The contributions of this study are threefold. Theoretically, the article enriches discourse on deep learning and comparative educational philosophy by incorporating two non-Western contexts, responding directly to the call for decolonial approaches (Hebert, 2023). Practically, the comparison provides a culturally nuanced reflective framework for educators, curriculum developers, and policymakers in both countries. From a policy standpoint, the article offers a conceptual foundation for the Indonesia-Uzbekistan educational partnership. The article proceeds as follows: an operational framework for deep learning, the research methodology, findings organised into three sub-themes, a discussion situating the findings within broader literature, research implications framed as testable propositions, limitations, and conclusion.

Conceptual Framework: An Operational Definition of Deep Learning

Because the comparison requires a common yardstick, deep learning is defined operationally before the analysis begins. Drawing on curriculum coherence scholarship (McPhail, 2021), critical conceptual analysis (Kovač et al., 2025), philosophical framing (Faizah et al., 2025), and the tripartite account of educational purpose advanced by Biesta (2015), this study treats deep learning as learning that satisfies four dimensions.

The first is epistemic depth, meaning engagement with the structure and internal logic of knowledge rather than its surface features, so that understanding can be transferred to unfamiliar situations. The second is learner agency, meaning that the learner exercises meaningful initiative over the direction, pace, and interpretation of learning, with intrinsic rather than externally imposed motivation. The third is moral and existential formation, meaning that learning shapes character, values, and the learner’s sense of self, and is not confined to the cognitive domain. The fourth is contextual and cultural embeddedness, meaning that learning is anchored in the learner’s lived social, cultural, and historical situation rather than delivered as decontextualised content. Table 1 sets out the four dimensions together with the textual indicators used to identify them during coding.

Two clarifications guard against anachronism. First, deep learning is a recent term in Indonesian educational policy, formalised in 2025, whereas Dewantara wrote in the middle of the twentieth century and the Uzbek instruments were drafted from 2019 onwards. Neither tradition used the term for itself. The framework is therefore applied as an external analytical lens for retrospective reading, in the same way that scholars read Confucian or Aristotelian texts through later categories without implying that those categories were available to the authors. Second, satisfying a dimension does not mean that a tradition anticipated contemporary policy. It means only that the texts contain principles which, read through this framework, support the practices deep learning now names.

Method

Research Design

This study uses a qualitative, comparative-philosophical design. The aim is to understand the meanings, values, and philosophical assumptions underlying two systems of educational thought from different cultural backgrounds, which calls for interpretive exploration of philosophical and policy texts rather than statistical generalisation (Creswell & Poth, 2018). The philosophical component analyses and reconstructs the core arguments of Dewantara’s philosophy and of Uzbekistan’s educational reform discourse. The comparative component identifies patterns of similarity and difference systematically, as is common in cross-cultural educational philosophy (Hebert, 2023).

Data Sources and the Named Primary Corpus

The primary technique is document study and textual analysis (Bowen, 2009). Table 2 lists the primary corpus on both sides. Documents were selected purposively, on the criterion that a text must state educational aims, principles of learning, or the cultural values underlying them, rather than address administrative or financing matters alone. This excluded a large volume of implementation regulations and budget instruments.

Two composition decisions require explanation. Permendikdasmen No. 13 of 2025 is included as a primary Indonesian instrument, because a philosophy-only Indonesian corpus set against a policy-only Uzbek corpus would have introduced a second and unacknowledged asymmetry. Including the regulation lets the Indonesian side be read at both the philosophical and the policy level. On the Uzbek side, the National Statement of Commitment submitted to the 2022 Transforming Education Summit was excluded, because it is a summit commitment rather than an instrument stating educational aims or principles of learning, and it therefore fails the criterion. It is treated as contextual material and is not cited as evidence. Secondary data comprise scholarly articles, books, and academic analyses on both topics, together with system-level reporting used for contextual description (OECD, 2024; UNICEF, 2017), giving access to the original thought of the figure and to official policy as primary sources (Mason, 2002).

Acknowledging Comparative Asymmetry

The two sides of this comparison are not of the same kind, and treating them as though they were would misrepresent both. Dewantara is a single thinker with a systematic framework. Uzbekistan’s educational reform thought is a diffuse policy discourse produced by state institutions over three decades, with no single equivalent figure. The asymmetry is a deliberate design choice rather than an oversight, and it is defensible on three grounds.

What is compared is philosophical foundations rather than authors, and assumptions about the person, knowledge, and the aims of education can be reconstructed from policy as legitimately as from authored work. The asymmetry is itself part of the finding, since the contrast between an individual voice and a collective state discourse mirrors the divergence between bottom-up and top-down orientations that the analysis identifies. And the Uzbek corpus is not wholly impersonal, since policy mobilises Alisher Navoi, the fifteenth-century poet whose account of the perfect human being (komil inson) is invoked as the model of the educated person (Uzakova, 2025), alongside the recovery of pre-Soviet pedagogues such as Abdukadir Shakuri (Ashurova, 2025). Navoi functions as the closest available counterpart to Dewantara, though he is mobilised by the state rather than authoring the reform. Statements about the Uzbek side therefore describe an official philosophical orientation, not the settled conviction of a named philosopher.

Data Analysis

Analysis proceeded in four stages, adopting the thematic content analysis model of Braun and Clarke (2006) adapted for philosophical inquiry: repeated reading of each corpus separately to capture central ideas, deductive coding against the four dimensions in Table 1 supplemented by inductive codes for material the framework did not anticipate, cross-case comparison to identify convergence and divergence, and finally interpretation and synthesis.

Because coding philosophical prose differs from coding interview transcripts, an illustration is warranted. In the Indonesian corpus, Dewantara’s definition of education as guiding the natural potentials present in children so that they may attain the highest attainable safety and happiness as human beings and as members of society (Dewantara, 1967) was coded D2 on the strength of the verb menuntun, to guide, which positions the educator as one who accompanies growth already underway. The same passage was double-coded D3, because the stated end is the flourishing of the person, and D4, because that flourishing is specified as membership of a society. In the Uzbek corpus, Articles 3 and 19 of the Law on Education, defining lifelong learning and recognising adult education as formal, non-formal, and informal (Law No. ZRU-637, 2020), were coded D2 for their recognition of learning outside institutional direction. The same law was not coded D2 where it assigns curriculum standards to state bodies, and that material received an inductive code for centralised determination. Codes were retained only where the textual indicator was explicit, and inferred meanings were recorded separately in the audit trail.

Table 4 reports the resulting evidence map at passage level, so that dimension coverage is visible rather than implied. Twenty passages were retained, eleven from the Indonesian corpus and nine from the Uzbek, distributed evenly across the four dimensions at five each, but with the Indonesian and Uzbek contributions to D1 differing in kind rather than in count. That difference is reported as a finding in its own right below.

Trustworthiness

Several strategies were applied. Source triangulation compared primary and secondary documents within each context. The Uzbek legal and strategic instruments were read in the original Uzbek and Russian by the second author and checked against the published English translations where these exist, so that no part of the analysis rests on translation alone. An audit trail documented decisions from data collection through to conclusions, including the coded status of every passage cited in the findings, and Table 4 makes that record public rather than internal. The team spans the two traditions under comparison, since the first author is positioned within the Indonesian tradition and the second within the Uzbek. That position supports interpretation on each side, though it also carries the risk that either author reads the unfamiliar corpus through a familiar frame, and both authors engaged in continuous self-reflection to limit it. Historical and cultural context was explained to enable transferability to similar dialogue contexts (Creswell & Poth, 2018). Coding was nonetheless carried out by the first author alone, with no second coder and no intercoder agreement statistic, even though the second author supplied and verified the Uzbek source material. This is disclosed rather than mitigated by claim, and the passage-level display in Table 4 is offered as the practical substitute, since it lets a reader test each coding decision against the source text directly.

Findings

Core Principles and Their Relevance to Deep Learning

Analysis of Ki Hadjar Dewantara’s educational philosophy reveals principles that form an educational ecosystem supportive of deep learning across all four dimensions. Dewantara defines the purpose of education as guiding the natural forces present in children so that they may reach, as human beings and as members of society, the highest attainable safety and happiness (Dewantara, 1967). Three features of this formulation carry analytical weight. The verb of guidance rather than instruction places the educator behind the learner’s own movement, satisfying D2. The dual address to the person and to the member of society satisfies D3 and D4 simultaneously. The absence of any reference to the transmission of content marks the definition as one about formation rather than delivery.

Epistemic depth is evidenced most directly in Dewantara’s Tri-Nga, the triad of ngerti, ngrasa, and nglakoni, meaning to understand, to feel, and to enact (Dewantara, 1967). The triad is coded D1 because it specifies a criterion of adequacy for knowing: understanding that is not felt and not enacted is treated as incomplete rather than as partial. Knowledge is held to be complete only when it has been carried into conduct, which is a claim about transfer stated in normative rather than psychological terms. The reception literature supports this reading, describing the triad as an insistence on balance between cognitive, affective, and enacted registers rather than on intellectual training alone (Darmawan & Sujoko, 2019; Siswadi & Murtiningsih, 2024).

The principle of kemerdekaan (freedom) in Dewantara’s thought is not boundless liberty but the creation of conditions under which the learner’s kodrat alam (natural disposition), meaning their talents and interests, may develop organically (Dewantara, 1967; Basuni et al., 2021). Its operationalisation is realised through the Sistem Among, in which the teacher acts as a pamong or guide. The pedagogical triad “ing ngarsa sung tuladha, ing madya mangun karsa, tut wuri handayani”, meaning to set an example when in front, to build will when in the midst, and to give encouragement from behind, specifies three distinct teacher positions rather than one (Efendy, 2023; Suwignyo, 2021). Read against Table 1, the triad is significant because only the first position involves modelling and none involves transmission. The teacher paradigm shifts from source of knowledge to facilitator of contextual and meaningful experience, a condition McPhail (2021) treats as necessary for coherent understanding. Dewantara’s epistemology has accordingly been characterised as anti-positivist and open to plural meaning (Ramadan, 2025), and his framework positioned as a systematic philosophy of education rather than a set of maxims (Siswadi & Murtiningsih, 2024; Tilaar, 2002).

Dewantara’s cultural writings supply the fourth dimension and reinforce the first. The Trikon principle holds that cultural and educational development must be kontinu, continuous with a community’s own inheritance; konvergen, open to convergence with other cultures; and konsentris, concentric, in that a community joins the universal while retaining its own circle (Dewantara, 1977). The passage was coded D4 for the explicit anchoring of learning in cultural inheritance, and additionally D1 for the convergence element, which requires that knowledge drawn from elsewhere be integrated into an existing structure rather than accumulated beside it. The commitment is coherentist rather than eclectic, which is precisely the property McPhail (2021) identifies as the condition of depth. The same principle underwrites the concept of Tri Pusat Pendidikan, the three centres of education comprising family, school, and youth community, which locates learning in life rather than in the classroom alone.

This philosophical account now has an explicit policy counterpart. Permendikdasmen No. 13 of 2025 mandates deep learning across primary and secondary schooling from the 2025/2026 academic year, and frames it as an approach in which students do not merely memorise material but understand it as a whole, connect concepts to one another, and apply them in different contexts (Permendikdasmen No. 13, 2025). The instrument was coded D1 on that formulation and D4 on its provision for linking topics to cultural, social, and everyday contexts across subject boundaries. Its inclusion matters analytically because it shows the Indonesian side of the comparison expressing the same dimension at two levels, philosophical and regulatory, with the regulatory formulation reading as a policy restatement of what Tri-Nga and Trikon had already articulated.

The philosophical foundation of Uzbekistan’s post-independence educational reform contextualises the same dimensions within nation-building and cultural identity revitalisation. The Law on Education (No. ZRU-637, 2020) reconstitutes the education system as a continuous whole. In recognising adult education as a central component of lifelong learning that includes formal, non-formal, and informal processes, it extends the reach of learning beyond institutional walls in a manner consistent with D2 and D4. The same instrument protects learners from corporal and psychological violence in educational organisations, a provision that carries philosophical weight because it presumes the learner as a bearer of dignity rather than an object of formation, and it was coded D3 on that basis. At the level of strategy, the Development Strategy of New Uzbekistan for 2022-2026 was framed under the principle of moving from a strategy of actions to a strategy of development (Presidential Decree No. PF-60, 2022), signalling a shift from discrete measures towards systemic transformation.

Epistemic depth on the Uzbek side is present but expressed differently, and the difference is worth stating plainly rather than smoothing over. The Concept for the Development of the Higher Education System until 2030 commits the system to a phased transfer of the educational process to a credit-module structure, alongside targets to raise higher education coverage above 50 percent and to place at least ten institutions within the first thousand positions of internationally recognised rankings (Presidential Decree No. PF-5847, 2019). The credit-module provision was coded D1, since modularisation is a decision about how knowledge is segmented, sequenced, and accumulated, and therefore a claim about the structure of knowledge rather than about its delivery. The coverage and ranking targets in the same instrument were not coded D1. They were assigned an inductive code for instrumental and competitive rationale, because they specify the scale and external validation of provision rather than the character of understanding. The “Uzbekistan - 2030” Strategy repeats this pattern, setting quantified goals for coverage, international rankings, and joint programmes with foreign universities (Presidential Decree No. PF-158, 2023). The Uzbek corpus therefore satisfies D1 through system architecture and international benchmarking rather than through an account of what it is to understand something deeply. That is a real difference from the Indonesian side, and it is reported here as a finding rather than absorbed into a claim of equivalence.

Reform is intended not only to improve technical skills but also to shape a generation with strong spiritual and moral roots, an aim documented across the policy-facing literature (Sharipov, 2022; Akhmedova, 2024) and corroborated by indexed scholarship on moral and spiritual formation in Central Asian education (Vinogradova et al., 2025). This is D3 stated in national rather than individual terms. Deep learning in Uzbekistan can accordingly be understood as a process of internalising national and religious values, reflected in efforts to integrate the intellectual heritage of thinkers such as Alisher Navoi into the curriculum (Uzakova, 2025) and in the recovery of pre-Soviet reformist pedagogy (Ashurova, 2025; Kundizalieva, 2023). This focus on character finds common ground with Dewantara’s emphasis on budi pekerti (moral character), although the values developed stem from different cultural roots. The principal driver of change diverges sharply. Dewantara emphasised the individual’s liberation from colonial constraint. Uzbek reform is driven by a post-Soviet national agenda to forge a new collective identity, a reading supported by political-economy analysis of education in Central Asia (Ozawa et al., 2024).

Convergence and Divergence in Educational Aims and Processes

The comparative analysis identifies convergence in the view of educational purpose. Both Dewantara and the discourse of Uzbek reform reject education that is merely instrumental and confined to the cognitive dimension. Both emphasise formation of the whole person, academically capable, of strong character, and able to contribute to society (Budiono, 2017; Vinogradova et al., 2025). The concept of Tri Pusat Pendidikan advanced by Dewantara, encompassing the family, the school, and the youth community, emphasises learning from and within life. It resonates with the Uzbek reconstitution of education as a continuous system spanning preschool through adult learning (Law No. ZRU-637, 2020; Mamadaliyeva & Qosimova, 2024), though the resonance is structural rather than genealogical.

Divergence emerges in practical implications and in the emphasis placed on the educational process. Through the principle of kodrat alam, Dewantara places the individuality of the learner as the starting point, so that the educational process follows the rhythm of the child’s development rather than the reverse (Darmawan & Sujoko, 2019). Within the macro-level narrative of Uzbek reform there is greater pressure towards collective national objectives, including global competitiveness and unified national identity (Gafurov, 2024; Mavlonov & Bo’riyev, 2025; Presidential Decree No. PF-158, 2023). The state’s role in determining curriculum direction and educational standards is more central than in Dewantara’s philosophy, which is decentralised and community-based. Deep learning from Dewantara’s perspective tends towards the creation of personal meaning. In the Uzbek context it is directed more towards mastery of the competencies and values the state identifies as necessary. Table 3 sets out the comparison.

The final row states the equifinality noted in the introduction. The two traditions arrive at broadly similar commitments about what learning should be while distributing authority over learning in opposite directions. This is the result that neither case yields alone. Figure 1 represents the relation.

Figure 1. Two institutional routes converging on the four dimensions of deep learning

Source: Author’s construction based on the analysis in Tables 3 and 4.

Evidence Base and Dimension Coverage

Because the findings above are organised by tradition and then by convergence and divergence, dimension coverage is not otherwise visible to the reader. Table 4 therefore maps each dimension to the specific passages coded in each corpus. It serves three purposes. It shows that every document named in Table 2 generates evidence used in the analysis. It exposes the distribution of evidence across dimensions rather than leaving it implied. And it makes the imbalance on D1 legible, since the Indonesian entries for that dimension are philosophical statements about the completeness of knowing while the Uzbek entries are structural provisions about how knowledge is organised and benchmarked.

Two observations follow from the table. First, D3 and D4 are the best-evidenced dimensions on both sides, which is unsurprising given that both traditions are explicit about moral formation and cultural anchoring. Second, D1 is the dimension on which the corpora diverge most sharply in kind. The Indonesian entries are statements about what it means to know something; the Uzbek entries are statements about how a system arranges and validates knowing. The two are not equivalent, and the comparison is stronger for saying so. The equifinality claim advanced below is a claim about the dimensions being reached, not about their being reached in the same idiom.

Discussion

The findings reinforce the thesis that deep learning is not a culturally neutral concept. Effective implementation requires alignment with, and reinterpretation of, philosophical values already present in a society (Faizah et al., 2025; Hebert, 2023). Both Dewantara’s philosophy and the Uzbek reform framework offer grounds for constructing an authentic meaning of deep learning, with different emphases. Dewantara’s thought is child-centred. The Uzbek framework is society-centred and nation-centred. This extends the critical analysis of Kovač et al. (2025) regarding cultural bias in deep learning discourse by demonstrating that non-Western frameworks possess their own conceptualisations of deep and meaningful learning.

The convergence in educational aims indicates common ground for dialogue. It aligns with Biesta’s (2015) tripartite framework of qualification, socialisation, and subjectification, given that both traditions prioritise socialisation and subjectification alongside cognitive development. This convergence matters for resisting the reduction of deep learning to a checklist of twenty-first-century competencies, a concern raised by Kovač et al. (2025). Both traditions demonstrate that authentic deep learning encompasses moral, spiritual, and cultural dimensions beyond instrumental skill.

The divergence between individual and collective emphasis, and between bottom-up and top-down routes, need not be treated as an obstacle. The tension has the potential to enrich cooperation. Indonesia can remind Uzbekistan of the importance of preserving space for individual freedom within large-scale reform. Uzbekistan can offer Indonesia a perspective on aligning educational aims with a clear vision of national development. This dialectic resonates with debates in comparative education concerning the balance between individual development and collective welfare (Ryan & Louie, 2007), and between global frameworks and local adaptation (Sukjairungwattana et al., 2025). Ryan and Louie’s warning against false dichotomy is directly applicable here, since the equifinality identified in Table 3 shows that the individual and collective routes are not mutually exclusive paths to depth.

The equifinality finding also bears on implementation research. Studies of teachers moving towards deep learning pedagogy show that the shift is mediated by the structures within which teachers work rather than determined by the concept itself (Lim et al., 2016), and early Indonesian implementation reviews report the same dependence on institutional conditions (Feri et al., 2025). If depth can be reached through either a devolved or a directed institutional route, then the relevant question for policy is not which philosophy is more conducive to deep learning but whether a given system supplies the conditions its own philosophy requires. The uneven evidence for D1 sharpens the point rather than weakening it. A system that secures depth through modularisation and benchmarking, as the Uzbek instruments do, is relying on structure to produce a quality of understanding that its own texts do not describe. Whether structure alone is sufficient for that purpose is an empirical question the present design cannot settle.

The study also responds to the call for decolonial approaches in comparative educational philosophy (Hebert, 2023). By positioning two non-Western traditions as the primary subjects of analysis, it contributes to diversifying global discourse on deep learning and quality education. A note on one adjacent source is warranted here, since the article rests on keeping the pedagogical and computational senses of the term apart. Zhang (2020) applies human comprehensive development theory to innovation education and evaluates students’ deep learning status through an evaluation model, so the article uses the term in its pedagogical sense while operationalising it through computational modelling. It is cited here not as support for the pedagogical reading but as an illustration of how readily the two senses are combined in the literature, which is precisely the ambiguity the present article seeks to avoid.

Research Implications

The implications of a philosophical comparison are prospective, so what follows should be read as propositions for future empirical work rather than as findings. They arise from the analysis but have not been tested by it.

Theoretically, the study expands and decolonises deep learning discourse by rooting it in two non-Western philosophical systems. Meaningful engagement, knowledge construction, and character formation were already embedded in Dewantara’s thought and in Uzbek reform before the term became globally current, which challenges its standing as a purely Western and technocratic concept and answers Hebert’s (2023) call for dialogue between Asian intellectual traditions. The equifinality result carries a further theoretical weight. If the same four dimensions can be reached through opposite institutional routes, then arguments that tie deep learning necessarily to learner-centred devolution are underdetermined by the evidence.

Practically, four claims follow for Indonesia-Uzbekistan cooperation, each stated so that it can be tested. Bilateral dialogue conducted at the level of philosophical paradigm rather than programme exchange alone should produce agreements with greater longevity and lower renegotiation cost, which the implementation records of existing education agreements would reveal. Student exchange built around joint inquiry into the two traditions, examining tut wuri handayani alongside the Turkic tekke tradition of spiritual learning (Egamberdiyev & Azmukhanova, 2025), should yield larger intercultural gains than exchange organised around subject study, measurable across matched cohorts before and after the programme. Joint curriculum modules that integrate character and local wisdom from both nations, including Indonesian language teaching in Uzbekistan, should raise learner engagement above modules confined to linguistic form, a question comparative module trials could settle and a priority both ministries have identified (Kemendikdasmen, 2025). Teacher training that develops reflective competence regarding the philosophical foundations of practice (Mustaeva, 2026) should produce more durable pedagogical change than training focused on technique, and existing evidence supports this claim most strongly (OECD, 2024; Lim et al., 2016).

Research Limitations

Several limitations should be acknowledged. The study relies on textual analysis, which cannot show how these philosophies are interpreted or distorted in classroom practice; observation and interviews would add an empirical dimension. The comparison is asymmetric by design, a single thinker set against a state policy discourse. Navoi is mobilised by policy rather than its author, so claims about the Uzbek philosophical position are claims about an official orientation. Coding was carried out by the first author alone, with no second coder and no intercoder agreement statistic, and the division of labour between the authors was by corpus rather than by analytical stage. Table 4 lets coding decisions be checked against the sources, but a second coder would strengthen confidence. Evidence for D1 differs in kind between the two corpora, so the equifinality claim holds more firmly for D2, D3, and D4. The Uzbek primary instruments were read in the original, but the secondary literature consulted was predominantly English-language, so parts of the Uzbek- and Russian-language scholarly debate may be under-represented. The “Uzbekistan-2030” Strategy was updated in February 2026, so readers should consult the current consolidated text. The cooperation propositions remain conceptual and untested.

Conclusion

Deep learning, read through a comparative lens rarely applied, proves present in both Ki Hadjar Dewantara’s educational philosophy and the philosophical foundations of Uzbekistan’s educational reform, though each tradition expresses contextual understanding, character formation, and liberating pedagogy through its own cultural idiom and historical situation. Dewantara grounds learning in the child through the Sistem Among, with Tri-Nga supplying a criterion for when knowing is complete and Trikon setting the terms on which external knowledge is absorbed; Uzbekistan frames it within nation-building and the revitalisation of cultural identity. Both reject reductive and instrumental education, yet they differ in the weight given to individual and collective orientation and, more consequentially, in the opposite institutional routes by which depth is delivered, a divergence that must be read against uneven evidence for epistemic depth, stated philosophically in the Indonesian corpus and structurally in the Uzbek one. Dialogue between these foundations offers a basis for moving Indonesia-Uzbekistan educational cooperation beyond technical exchange towards a sustained strategic partnership, and for pursuing the shared aim of an education that produces intellectually capable individuals who are also whole persons of strong character. Field studies with educators and students in both countries, wider engagement with Uzbek- and Russian-language scholarship, and extension of this approach to other South-South partnerships would test the propositions advanced here.

Declarations

Funding. This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Author Contributions. Fridiyanto conceived the study, developed the four-dimension analytical framework, assembled and analysed the Indonesian corpus, carried out the coding reported in Table 4, and drafted the manuscript. Khakimova assembled the Uzbek corpus, read and verified the Uzbek- and Russian-language primary texts, contributed the interpretation of Uzbek reform discourse, and reviewed the manuscript for accuracy on the Uzbek side. Both authors read and approved the final version.

Ethical Approval. Ethical approval was not required for this study. The research involved no human participants, no animal subjects, and no personally identifiable data, and it draws exclusively on published scholarly works and publicly accessible legal and policy documents.

Use of Artificial Intelligence. The authors declare the use of Claude (Anthropic) for language editing and reference formatting only. No part of the analysis, interpretation, or conclusions of this study was generated by the tool. All output was reviewed by the authors, who take full responsibility for the content of the manuscript.

References

1. Akhmedova, D. K. (2024). Social and philosophical analysis of the reforms in the New Uzbekistan education system. Buletin Antropologi Indonesia, 1(3). https://doi.org/10.47134/bai.v1i3.3217
2. Ashurova, D. I. (2025). Types of education in the works of Abdukadir Shakuri and their significance today. European International Journal of Pedagogics, 5(3), 34-37. https://doi.org/10.55640/eijp-05-03-10
3. Azhar Fuadi, D., Yuniarti, G., Nelesti, N., Heri, S., Carnolis, J., Sri Hartati, M., & Susiyanto, S. (2025). Ki Hadjar Dewantara and the philosophy of education. Indonesian Journal of Innovation Multidisipliner Research, 3(1), 430-441. https://doi.org/10.69693/ijim.v3i1.324
4. Badan Pengembangan dan Pembinaan Bahasa. (2024). Peta bahasa: Data bahasa daerah di Indonesia. Kementerian Pendidikan Dasar dan Menengah. https://petabahasa.kemendikdasmen.go.id
5. Basuni, A., Ma’mur, I., & Kurniawan, A. F. (2021). Freedom to learn manifesto of Ki Hadjar Dewantara’s educational philosophy (A new formula in PAI learning). International Journal of Nusantara Islam, 9(2), 548-560. https://doi.org/10.15575/ijni.v9i2.16871
6. Biesta, G. (2015). What is education for? On good education, teacher judgement, and educational professionalism. European Journal of Education, 50(1), 75-87. https://doi.org/10.1111/ejed.12109
7. Bowen, G. A. (2009). Document analysis as a qualitative research method. Qualitative Research Journal, 9(2), 27-40. https://doi.org/10.3316/QRJ0902027
8. Braun, V., & Clarke, V. (2006). Using thematic analysis in psychology. Qualitative Research in Psychology, 3(2), 77-101. https://doi.org/10.1191/1478088706qp063oa
9. Budi Wijaya, I. K. W., Atmadja, A. T., Suastra, I. W., & Tika, I. N. (2024). Implementasi filsafat pendidikan modern dan filsafat pendidikan Ki Hadjar Dewantara dalam pembelajaran kimia. SCIENCE: Jurnal Inovasi Pendidikan Matematika dan IPA, 4(4), 504-510. https://doi.org/10.51878/science.v4i4.3531
10. Budiono, B. (2017). Pendidikan humanistik Ki Hajar Dewantara dalam perspektif pendidikan Islam. Intelektual: Jurnal Pendidikan dan Studi Keislaman, 7(1), 42-53. https://doi.org/10.33367/intelektual.v7i1.360
11. Creswell, J. W., & Poth, C. N. (2018). Qualitative inquiry and research design: Choosing among five approaches (4th ed.). Sage Publications.
12. Darmawan, I. P. A., & Sujoko, E. (2019). Understanding Ki Hadjar Dewantara’s educational philosophy. International Journal of Humanities and Innovation, 2(3), 65-68. https://doi.org/10.33750/ijhi.v2i3.42
13. Dewantara, K. H. (1967). Bagian pertama: Pendidikan. Majelis Luhur Persatuan Taman Siswa.
14. Dewantara, K. H. (1977). Kebudayaan. Majelis Luhur Persatuan Taman Siswa.
15. Efendy, T. (2023). Konsep sistem among dalam pendidikan menurut Ki Hadjar Dewantara. Jurnal Multidisiplin Indonesia, 2(6), 1105-1112. https://doi.org/10.58344/jmi.v2i6.274
16. Egamberdiyev, M., & Azmukhanova, A. (2025). The phenomenon of “Tekke” in the context of the spiritual and cultural identity of Turkic-Muslim peoples (Mid 19th-Early 20th century). Gumilyov Journal of History, 150(1), 28-45. https://doi.org/10.32523/3080-129x-2025-150-1-2-28-45
17. Faizah, H., Hendri, G., & Andriana, S. (2025). Deep learning in the perspective of educational philosophy: Between existence, ethics, and epistemology. International Journal of Educational Research, 2(4), 1-9. https://doi.org/10.62951/ijer.v2i4.430
18. Feri, M., Ismiati, N., Al-Nur, W. R., & Akbar, F. N. (2025). Implementing deep learning approaches in primary education: A literature review. Jurnal Varidika, 37(2), 162-171. https://doi.org/10.23917/varidika.v37i2.12151
19. Gafurov, A. (2024). Reforms in the education system of the Republic of Uzbekistan to improve the management of higher education institutions. Budi Luhur Journal of Strategic and Global Studies, 2(1), 80-95. https://doi.org/10.36080/jsgs.v2i1.27
20. Hebert, D. G. (2023). Why comparative and decolonial studies in philosophy of education? In D. G. Hebert (Ed.), Comparative and decolonial studies in philosophy of education (pp. 1-13). Springer. https://doi.org/10.1007/978-981-99-0139-5_1
21. Kemendikdasmen. (2025, November 10). Siaran pers: Mendikdasmen bahas pendekatan pembelajaran mendalam dan pendidikan karakter di Uzbekistan. Kementerian Pendidikan Dasar dan Menengah Republik Indonesia. https://kemendikdasmen.go.id/siaran-pers/14081-mendikdasmen-bahas-pendekatan-pembelajaran-mendalam-dan-pendidikan-karakter-di-uzbekistan
22. Kemendikdasmen. (2025b). Data pokok pendidikan: Rekapitulasi nasional peserta didik. Kementerian Pendidikan Dasar dan Menengah Republik Indonesia. https://dapo.kemendikdasmen.go.id/pd
23. Kovač, V. B., Nome, D. Ø., Jensen, A. R., & Skreland, L. Lj. (2025). The why, what and how of deep learning: Critical analysis and additional concerns. Education Inquiry, 16(2), 237-253. https://doi.org/10.1080/20004508.2023.2194502
24. Kundizalieva, S. K. (2023). Philosophy of education: Theoretical conceptual basis for the development of the educational process. World Bulletin of Social Sciences, 20, 148-151. https://www.scholarexpress.net/index.php/wbss/article/view/2418
25. Law of the Republic of Uzbekistan No. ZRU-637. (2020, September 23). On education (new version). National Database of Legislative Information. https://lex.uz/ru/docs/5700831
26. Lim, W. Y., Ong, A., Soh, L. L., & Sufi, A. (2016). Teachers’ voices and change: The structure and agency dialectics that shaped teachers’ pedagogy toward deep learning. In C. S. Chai, C. P. Lim, & C. M. Tan (Eds.), Future learning in primary schools (pp. 147-158). Springer. https://doi.org/10.1007/978-981-287-579-2_10
27. Mamadaliyeva, Z., & Qosimova, M. (2024). Education system of Uzbekistan. Journal of Education, Ethics and Value, 3(11), 41-44. https://jeev.innovascience.uz/index.php/jeev/article/view/920
28. Mason, J. (2002). Qualitative researching (2nd ed.). Sage Publications.
29. Mavlonov, J. Y., & Bo’riyev, I. S. (2025). Educational policy in Uzbekistan: Experience and modernization processes. International Journal of Scientific Trends, 4(1), 114-121. https://scientifictrends.org/index.php/ijst/article/view/189
30. McPhail, G. (2021). The search for deep learning: A curriculum coherence model. Journal of Curriculum Studies, 53(4), 420-434. https://doi.org/10.1080/00220272.2020.1748231
31. Mustaeva, E. Y. (2026). Development of reflective competence in in-service teachers as a pressing pedagogical issue. Current Research Journal of Pedagogics, 7(1), 33-36. https://doi.org/10.37547/pedagogics-crjp-07-01-07
32. OECD. (2024). Results from TALIS 2024 country notes: Uzbekistan. OECD Publishing. https://www.oecd.org/en/publications/results-from-talis-2024-country-notes_e127f9e2-en/uzbekistan_96cdef5e-en.html
33. Ozawa, V., Durrani, N., & Thibault, H. (2024). The political economy of education in Central Asia: Exploring the fault lines of social cohesion. Globalisation, Societies and Education, 1-14. https://doi.org/10.1080/14767724.2024.2330361
34. Permendikdasmen No. 10. (2025). Peraturan Menteri Pendidikan Dasar dan Menengah Nomor 10 Tahun 2025 tentang Standar Kompetensi Lulusan. Kementerian Pendidikan Dasar dan Menengah Republik Indonesia.
35. Permendikdasmen No. 12. (2025). Peraturan Menteri Pendidikan Dasar dan Menengah Nomor 12 Tahun 2025 tentang Standar Isi. Kementerian Pendidikan Dasar dan Menengah Republik Indonesia.
36. Permendikdasmen No. 13. (2025). Peraturan Menteri Pendidikan Dasar dan Menengah Nomor 13 Tahun 2025 tentang penyelenggaraan pembelajaran pada pendidikan dasar dan pendidikan menengah. Kementerian Pendidikan Dasar dan Menengah Republik Indonesia.
37. Prabtama, O. R., & Dwi Kartika, G. (2024). Ki Hajar Dewantara’s critique of the 2022 Sisdiknas Bill: Implications for Islamic education in Indonesia. Progresiva: Jurnal Pemikiran dan Pendidikan Islam, 13(3), 367-384. https://doi.org/10.22219/progresiva.v13i03.35589
38. Presidential Decree No. PF-60. (2022, January 28). On the Development Strategy of New Uzbekistan for 2022-2026. Republic of Uzbekistan. (Circulated as United Nations document A/76/684, 10 February 2022). https://digitallibrary.un.org/record/3959032
39. Presidential Decree No. PF-158. (2023, September 11). On the “Uzbekistan - 2030” Strategy. Republic of Uzbekistan. National Database of Legislative Information. (Updated by Presidential Decree No. PF-21, 16 February 2026). https://strategy.uz/index.php?static=strategy_actions&lang=en
40. Presidential Decree No. PF-5847. (2019, October 8). On approval of the Concept for the Development of the Higher Education System of the Republic of Uzbekistan until 2030. Republic of Uzbekistan. National Database of Legislative Information. English translation: https://cis-legislation.com/document.fwx?rgn=119515
41. Ramadan, R. (2025). Membongkar epistemologi pendidikan Ki Hadjar Dewantara dalam konteks postmodernisme. AN-NUR: Jurnal Pendidikan dan Pembelajaran, 1(2), 211-224. https://doi.org/10.65802/an-nur.v1i2.46
42. Ryan, J., & Louie, K. (2007). False dichotomy? “Western” and “Confucian” concepts of scholarship and learning. Educational Philosophy and Theory, 39(4), 404-417. https://doi.org/10.1111/j.1469-5812.2007.00347.x
43. Sharipov, A. Z. (2022). New Uzbekistan development strategy and national philosophy: Dialectic relationship and spiritual renewal. International Journal of Trend in Scientific Research and Development, (Special Issue), 44-48. https://www.ijtsrd.com/papers/ijtsrd52384.pdf
44. Siswadi, G. A., & Murtiningsih, R. S. (2024). Kedudukan pemikiran Ki Hadjar Dewantara dalam filsafat pendidikan. Padma Sari: Jurnal Ilmu Pendidikan, 4(1), 43-57. https://doi.org/10.53977/ps.v4i01.867
45. Sukjairungwattana, P., Hu, H., Liu, R., & Huang, J. (2025). From local to global: Systematically reviewing higher education internationalization in Asia. Frontiers in Education, 9, Article 1473820. https://doi.org/10.3389/feduc.2024.1473820
46. Suwignyo, H. (2021). Theoretical reconstruction for students’ critical writing competences based on Ki Hadjar Dewantara’s (KHD) Among System. In Proceedings of the International Seminar on Language, Education, and Culture (ISoLEC 2021) (pp. 127-134). Atlantis Press. https://doi.org/10.2991/assehr.k.211212.025
47. Tilaar, H. A. R. (2002). Pendidikan, kebudayaan, dan masyarakat madani Indonesia. Remaja Rosdakarya.
48. UNICEF. (2017, May). Policy review of preschool education in Uzbekistan. https://www.unicef.org/uzbekistan/media/746/file/Policy Review of Preschool Education.pdf
49. Uzakova, L. A. (2025). Alisher Navoi’s ideas in the education of a perfect generation: The experience of new Uzbekistan. Buletin Antropologi Indonesia, 2(2), 7. https://doi.org/10.47134/bai.v2i2.3888
50. Vinogradova, N. V., Bykov, S. V., Panov, I. G., Rotmistrov, V. I., & Zuev, A. V. (2025). Socio-humanitarian aspects of the problem of development of spiritual and moral consciousness of contemporary society: Means of education. In E. G. Popkova (Ed.), Lifelong learning in Central Asia (Education in the Asia-Pacific Region: Issues, Concerns and Prospects, Vol. 72, pp. 217-225). Springer. https://doi.org/10.1007/978-981-96-1941-2_21
51. Zhang, J.-L. (2020). The application of human comprehensive development theory and deep learning in innovation education in higher education. Frontiers in Psychology, 11, Article 1605. https://doi.org/10.3389/fpsyg.2020.01605