Neurocognitive Effects of Mandombe Writing Practice: A Cross-Sectional Study

Abstract

Title

From Grades to Phases: Learning Trajectories in Pilot Mandombe Nsandas of Kinshasa (2019–2023)

Ntima et al.

---

Abstract

This study examines how a pilot implementation of the new official Mandombe pedagogy alters the way learning trajectories should be described and assessed in Congolese primary schooling. We analyse longitudinal records from 2019–2023 in a subset of CENA’s Nsandas in Kinshasa that were designated as trial sites for the updated MEN-D curriculum and compare them with neighbouring schools following the standard national programme. The work tests, in a limited but concrete way, Nsiangani’s (2014) hypothesis that colonial age-grade assessment logic misrepresents the progress of children educated through indigenous symbolic systems, and that alternative phase-based units become necessary once Mandombe literacy is the base script.

For pilot Nsanda learners we reconstructed trajectories in two formats: (a) the conventional grade-per-year model (age, repetition, subject marks) and (b) a four-phase model derived from the official Mandombe pedagogy: Symbolic control (S), Geometric fluency (G), Transfer to mathematics and science (T), and Project/collective application (P). Each learner-year was independently coded to the highest phase consistently evidenced that year. For comparison schools we described trajectories in grade units only and explored whether an analogous phase coding was workable.

Under standard age-grade logic, pilot Nsanda trajectories appear irregular: reduced overage, frequent multi-grade promotion, and abrupt jumps in mathematics, literacy and project work. When the same data are replotted in phase units, a more coherent pattern emerges: many learners spend an extended period consolidating Mandombe symbolism (S), followed by faster movement through G, T and P, during which cross-domain gains cluster in a few years rather than spreading evenly across the cycle. Comparison schools show slower, more linear grade progression, higher overage, and fewer clustered leaps; attempts to apply the S–P phase rubric there proved unstable and conceptually forced.

We do not claim that all Nsandas behave identically, that pilot centres are universally superior, or that grades should be abolished. Our more modest conclusion is that in the Nsandas piloting the official Mandombe pedagogy, the conventional grade-per-year unit ceases to be the most informative descriptor of learning. A phase-based assessment logic, anchored in the internal structure of the curriculum, captures their trajectories more accurately and avoids misclassifying rapid, structurally organised progress as “irregular.” The protocol is simple and replicable so that other centres—including non-pilot Nsandas—can test whether similar phase signatures appear in their own data.

---

1. Introduction

1.1 Age-graded schooling as an industrial technology

The age-graded classroom, now taken for granted worldwide, is a relatively recent European invention. Until the late 18th century, most European schools were mixed-age, often organised in one room where older children informally assisted younger ones. Age-segregated grades emerged in German states and Prussia as part of wider reforms that linked mass schooling to state-building, military discipline and later to industrial needs: the efficient production of punctual, standardised workers and citizens.

As industrialisation advanced, age-graded, subject-segmented schools spread across Western Europe and North America, often displacing older monitorial or community-based models. European empires then exported this organisational form to their colonies, where it served both as a symbol of “modernity” and as an instrument for producing a disciplined, low-autonomy workforce. In the Congo, the colonial school was explicitly designed around these logics, with indigenous languages and symbolic systems displaced in favour of French and imported curricula. The age-grade grid thus arrived not as a neutral pedagogical innovation, but as part of a broader industrial-colonial package.

1.2 Why this matters for Mandombe

The MEN-D Mandombe programme re-centres an indigenous script as the primary literacy and cognitive scaffold. In the new official pedagogy, letters are geometric objects; reading and writing involve mastery of rotation, symmetry and branch logic; and these structures are progressively extended to mathematics, science and collective projects. In such a curriculum, the unit that organises content and expectations is no longer simply “one grade per year” but the learner’s position in a series of symbolic-cognitive phases.

Nsiangani (2014) argued that evaluating such programmes solely through the inherited age-grade logic risks systematic distortion. Where a curriculum is designed around phases (long installation of a powerful symbolic system, then rapid cross-domain leverage), an assessment grid that assumes linear progression can misclassify successful trajectories as “irregular,” “unstable” or even “suspect.” This concern is not theoretical: pilot Nsandas were often asked by inspectors to explain why some learners “skipped” grades or completed cycles “too quickly,” as if compression were evidence of manipulation rather than a design feature of the pedagogy.

1.3 Focus of this study

This paper concentrates on a subset of Nsandas in Kinshasa that were officially designated as pilot sites for the updated Mandombe pedagogy and assessment tools. Other Nsandas continued with earlier variants of the programme or hybrid models; we do not assume their trajectories match those described here.

Using only routine records from 2019–2023, we ask three questions:

1. How do learners in pilot Nsandas appear when described purely through the colonial-industrial age-grade logic (age, grade, repetition, annual marks)?

2. Can the same data be re-described through a small set of curriculum-anchored learning phases that better reflect the internal structure of the official Mandombe pedagogy?

3. Do these phase-based descriptions differ in systematic ways from trajectories observed in neighbouring non-Mandombe schools that operate entirely within the inherited age-grade model?

The aim is modest and empirical: to see whether the choice of assessment unit (grade vs phase) materially alters how the pilot centres’ work is perceived and whether a phase logic is operationally usable without heroic data requirements.

---

2. Conceptual framework

2.1 Colonial age-grade logic revisited

Age-grading combines three elements: chronological age, the notion of a fixed curriculum slice per year, and promotion rules that make the grade the primary indicator of progress. Historians link its consolidation to 18th–19th century German/Prussian reforms and to 19th-century industrial rationalisation, where factories and bureaucracies required cohorts of workers trained to similar specifications at each step.

In African colonies, age-graded schools were introduced as part of a “civilising” mission, but they also served labour and administrative needs: they produced clerks, interpreters and low-level technicians who were literate in the coloniser’s language and accustomed to hierarchical obedience. When independence came, many countries, including the DRC, retained this model with only limited adaptation. The age-grade ladder thus remains not only a technical device but a historical trace of industrial-colonial priorities.

2.2 Phases in the official Mandombe pedagogy

The new Mandombe pedagogy, as validated by the DRC Ministries of Education and of Higher Education & Scientific Research, structures learning around overlapping phases rather than calendar years. For this study, and in consultation with trainers, we summarise four observable phases:

S – Symbolic control: secure decoding and production of basic Mandombe units, with attention to mvuala, branches, and orientation; frequent teacher scaffolding.

G – Geometric fluency: spontaneous manipulation of rotations and mirrorings; use of Mandombe forms in games and self-initiated drawings; internal assessments at mastery level for geometric tasks.

T – Transfer to maths/science: explicit use of Mandombe structures to reason about number, operations and simple physical relations (e.g., branch-based grouping, symmetry in equations).

P – Project/collective application: integration of Mandombe and mathematics in multi-step group projects; documented initiative, explanation to peers, and problem-solving.

The phases are not tied to age; a late-starting 11-year-old and a 7-year-old may both be in Phase S, and the older learner may move through G/T/P faster once symbolic control is achieved.

2.3 Hypotheses

We formulate three cautious hypotheses for the pilot Nsandas:

H1 – Descriptive mismatch. When plotted purely in age-grade terms, pilot Nsanda trajectories will appear irregular (compressed cycles, multi-grade promotion, non-linear marks).

H2 – Phase coherence. Recoding the same data in S–P phases will reveal more regular trajectories (longer S, then faster G/T/P) with cross-domain gains clustering around phase transitions.

H3 – Programme specificity. Attempts to apply the S–P rubric to neighbouring non-Mandombe schools, which follow the inherited industrial age-grade model without Mandombe, will prove less reliable and less informative, indicating that the phase logic is anchored in the Mandombe pedagogy rather than in generic schooling.

If any of these fail—for instance, if grade-logic already describes pilot Nsanda trajectories well, or if phase coding is noisy and uninformative—then the case for revising assessment logic weakens.

---

3. Method

3.1 Setting: pilot Nsandas and comparison schools

The study covers several Nsandas in Kinshasa formally designated by the MEN-D steering committee as pilot sites for the new official Mandombe pedagogy and internal assessment tools. These centres applied the full maternal-language + Mandombe curriculum from early primary and received additional supervision and documentation support.

For comparison, we selected nearby public or low-fee private schools using the standard national curriculum in French and Latin script, with conventional age-graded classes and no Mandombe instruction. These schools represent the industrial-colonial model in its current Congolese form.

3.2 Participants and inclusion

Within pilot Nsandas and comparison schools, we included learners who:

had at least three consecutive learner-years recorded between 2019 and 2023;

remained in the same school during that interval;

had complete annual records (age, grade, core subject marks, teacher comments).

This yielded several hundred learner-years across sites. The sample is thus institutionally focused (on centres), not nationally representative.

3.3 Data and coding

For each learner-year, we extracted:

Administrative data: age, official grade, promotion/repetition, years over official age for grade;

Academic data: annual marks in language(s), mathematics, science and social studies;

Internal Mandombe data (pilot Nsandas only): scores on existing Mandombe mastery checks;

Teacher comments: qualitative notes on autonomy, project participation and difficulties.

3.3.1 Phase assignment in pilot Nsandas

Using curriculum documents and teacher interviews, we built an explicit rubric for assigning each learner-year to S, G, T or P based on documented evidence (work samples, assessments, comments). Two independent coders, blind to later outcomes, applied the rubric; inter-rater agreement was checked and disagreements resolved by consensus. Only the highest phase consistently supported by multiple indicators in that year was assigned.

3.3.2 Exploratory phase coding in comparison schools

To test H3, we attempted to apply an adapted version of the rubric to comparison schools, substituting generic indicators (e.g., use of diagrams in maths) where explicit Mandombe criteria were absent. This exercise was explicitly labelled exploratory; the expectation was that the rubric would map poorly onto data from schools that do not organise learning around the same symbolic phases.

3.4 Analytic strategy

Analyses proceeded in three steps:

1. Grade-logic description for pilot Nsandas and comparison schools: age-grade placement, overage, repetition, time to complete cycles, and mean marks by grade.

2. Phase-logic description for pilot Nsandas: time spent in each phase, sequences of S–P over calendar years, and alignment of phase transitions with changes in marks and comments.

3. Feasibility and stability of phase coding in comparison schools: agreement between coders, interpretability, and relation (or lack thereof) to changes in marks.

Given the institutional sample and reliance on routine data, the emphasis was on robust descriptive patterns rather than on fine p-values.

---

4. Results (summary)

4.1 Pilot Nsandas in grade units

Viewed purely through age-grade metrics, pilot Nsanda learners displayed:

substantially lower overage than peers in comparison schools;

compressed cycles, with some learners promoted across more than one grade over the period;

non-linear mark trajectories, with marked jumps in mathematics and general average over one or two years.

From a conventional inspector’s perspective, these patterns can appear suspicious or unstable, as they depart from industrial expectations of steady, grade-by-grade progress.

4.2 Pilot Nsandas in phase units

Replotting the same learners in S–P phase units produced more interpretable trajectories:

most spent a considerable initial period in Phase S, where grades changed slowly and marks rose modestly;

once evidence supported transition into Phase G, gains in mathematics and general reasoning often accelerated;

for learners reaching T and P, cross-domain gains (mathematics, science, project work) clustered within a short calendar window, explaining multi-grade promotions.

The apparent “irregularity” under grade logic thus corresponded to phase transitions designed into the Mandombe pedagogy: a long investment in symbolic installation, then rapid leverage.

4.3 Comparison schools

Comparison schools, operating under the inherited age-grade model without Mandombe, showed:

higher and more persistent overage;

fewer or no multi-grade promotions;

gradual, incremental improvements in marks without sharp clustering.

Attempts to apply the S–P rubric in these schools yielded lower coder agreement and ambiguous trajectories; phase labels did not align consistently with shifts in marks or comments. This supports the view that the phase framework is not a generic cognitive ladder but is specifically tied to the structure of the Mandombe curriculum.

---

5. Discussion

The findings from these pilot Nsandas suggest that once an indigenous symbolic system like Mandombe becomes the backbone of teaching, the inherited industrial-colonial age-grade logic no longer provides the most faithful description of learning. A phase-based logic, grounded in the official Mandombe pedagogy, aligns better with observed trajectories and helps avoid misreading rapid, structured progress as administrative irregularity.

We do not infer that all Nsandas in the DRC share this pattern; only a subset were piloting the new pedagogy during 2019–2023, and our data are limited to those centres. Nor do we treat phase coding as infallible: it remains partly interpretive and requires prospective refinement. The study nevertheless shows that with only routine school records, it is possible to operationalise an alternative assessment unit that respects the internal design of an indigenous curriculum and to put it into empirical conversation with the age-graded model inherited from European industrial history.

Future work should replicate the protocol in non-pilot Nsandas and other provinces, refine phase indicators, and examine how phase-based trajectories relate to external benchmarks such as national examinations and later academic or vocational success.