Ethical Dimensions of Symbolic Cognition: A Study on Responsibility and Collective Knowledge

Abstract

Ethical Dimensions of Symbolic Cognition:

Mandombe as a Relational Geometric Backbone for Teaching Responsibility

---

Abstract

We start from the hypothesis that Mandombe offers a relational geometric backbone which, when taught together with its traditional philosophy of interdependence, supports an ethical learning environment that a Latin-based script struggles to reproduce. In Mandombe, the base mvuala-shape is ambiguous and only becomes determinate when related to a kisimba whose orientation it also constrains; identity is thus resolved through mutually orienting relations rather than in isolation. We ask whether learners actually use this structure to think about relations and responsibility after only a brief introduction.

To test this, we ran a short, 3-hour module with novice learners in Kinshasa (N = 50), randomly assigned to a Mandombe condition or a Latin-based control. Both groups received the same mini-lesson on cooperation and responsibility; they differed only in the script they practiced (mvuala + kisimba vs a Latin letter and a boxed variant). At the end, all learners answered three neutral written questions comparing the two signs and imagining each as a person. Answers were coded for relational language and for explicit links between script structure and social or ethical meaning.

Mandombe learners more often described the lone mvuala as “not yet itself,” “missing something” or “unclear,” and used the addition of the kisimba to model support, belonging or shared responsibility. Latin learners, despite identical philosophical input, mostly gave literal or generic person descriptions and rarely grounded their answers in the visual structure of the signs. We interpret this as preliminary evidence that Mandombe’s geometry, aligned with its philosophy, provides a distinctive symbolic resource for teaching relational ethics. We also note an intriguing but unplanned contrast in how ambiguity and labelling were discussed, which we explicitly reserve for a separate, dedicated study.

(Note: the values in the Results are interpreted as examples of the expected pattern; despite the reasonable effect size, we did not aim to quantify the phenomenon but to detect it. We believe a more adapted dedicated protocol is needed to properly generalize. But that was not our aim here)

---

1. Introduction

Ethical reasoning is usually studied through language, explicit rules and individual choices. Symbolic systems, and writing systems in particular, tend to be treated as neutral tools that simply carry content. Yet scripts and notations are part of the environment that trains attention, categories and relations. They make some structures easy to see and rehearse, and leave others invisible.

In African contexts, this matters. Many antisocial patterns are not only individual traits but structural outcomes: they emerge from environments that normalise domination, dehumanisation and rigid labelling. In the DSM-H framework and related work on the “colonised mind” and Dark Tetrad leadership, scripts, languages and institutional symbols are treated as part of the pathology or the cure. They can either rehearse extractive, one-sided relations or support relational, reciprocal personhood.

Mandombe, a geometric script developed in Central Africa, is unusual because its internal logic explicitly resonates with African philosophies of relational personhood. In everyday practice, it is taught not only as a way to write but as a visualisation of societal values: identity through relation, balance rather than isolation, cohesion rather than fragmentation. Children do not just hear this; they enact it each time they draw a mvuala, attach a kisimba, and watch meaning appear from a relation.

This paper addresses a simple but important question. We ask whether, after a short and controlled introduction, novice learners actually use Mandombe’s geometry to think about relations and responsibility in ways that a Latin-based script does not support, even when the same ethical story is told. We do not claim behavioural or clinical outcomes. We do not claim that geometry “causes” ethics. Our goal is more modest and more basic: to show that Mandombe offers a relational geometric backbone that learners can mobilise as an ethical metaphor almost immediately, and that this affordance is not matched by a structurally simpler Latin script under comparable conditions.

We situate this in a broader programme. If such a backbone exists and can be reliably activated, it strengthens the case for treating Mandombe as part of an ethical and cognitive architecture aligned with African concepts of personhood, and provides a mechanistic bridge toward later DSM-H research on long-term transfer and possible buffering against antisocial structural pathologies.

---

2. Theoretical background

2.1 Mandombe geometry and relational identity

In Mandombe, the base mvuala-shape is ambiguous in isolation. As a pure shape it can correspond to one consonant or its 180° rotated counterpart. A child looking at the shape alone does not yet know “which” mvuala it is. Identity is resolved only when a kisimba (base frame) is attached. At the same time, the kisimba’s orientation is determined by the mvuala it defines.

Structurally:

mvuala-shape alone: underdetermined, ambiguous;

mvuala + kisimba: determinate consonant;

kisimba orientation: follows and reflects the mvuala.

The relation is mutual. The mvuala “needs” the kisimba for its identity, and the kisimba “takes its orientation” from the mvuala. Larger units are built by combining mvuala into mvuala mpamba, mvuala piluka and mvuala mpimpita, which gain stability and meaning when they meet other glyph classes. Identity is not an intrinsic property of isolated units. It is a relational resolution of ambiguity that appears when elements enter into structured relation.

Children practising Mandombe therefore spend time not only matching shapes to sounds, but enacting a logic of “alone / undefined” versus “with / clear,” of mutual orientation and stabilisation through combination. They see and reproduce, dozens of times per page, the idea that “what something is” depends on what it is with.

2.2 African philosophies of personhood

Afro-communitarian philosophies of personhood insist that a person is a person through other persons. Personhood is not a static attribute of an isolated individual but an achievement tied to participation, mutual recognition and responsibility in a community. In this view, it is incoherent to think of the self as a sealed, self-sufficient entity.

Relational personhood has ethical and ontological dimensions. Ethically, it foregrounds obligations and care: to be fully a person is to be responsible to others. Ontologically, it treats identity as emergent from webs of relation. These ideas sit uneasily with scripts and categories that are built for sharp, decontextualised labels and for ranking individuals in fixed, one-dimensional scales.

Mandombe, as taught in many centres, is explicitly linked to this relational ontology. Teachers often explain mvuala and kisimba using metaphors of family, groups, roles and support. A mvuala alone is presented as “not yet itself,” “not yet stable,” or “needing its kisimba.” Combinations are described as “cohesive units” that can carry meaning because they are balanced and in relation. The script is thus not only a code but a visual pedagogy of interdependence.

2.3 Symbolic affordances and ethical learning

We use the term ethical affordance to describe the ways in which a symbolic system makes certain ethical concepts easy or natural to express and rehearse. A script with a relational backbone, such as Mandombe, affords metaphors of support, mutual orientation and shared responsibility. A script based on atomic, context-free symbols affords quick labelling and categorisation.

We formulate two tiered hypotheses:

H₁ (structural): Because the mvuala-shape is ambiguous until attached to a kisimba whose orientation it constrains, Mandombe encodes identity as a relational resolution of ambiguity. Latin grapheme structure does not. This is a formal claim about the script, independent of any data.

H₂ (classroom): After a brief, balanced introduction in which both groups hear the same ethical story, learners working with Mandombe will more often use the structure of the script (alone vs with, ambiguous vs clarified, mutual orientation) to talk about relations and responsibility than learners working with a Latin-based script.

We deliberately leave clinical and long-term neuroplastic hypotheses (H₃) for future work. This paper focuses on H₂: classroom-level evidence that the relational backbone is not only present in theory but activated in novice practice.

---

3. Method

3.1 Setting and participants

The study took place in two urban schools in Kinshasa. We recruited 50 learners (age range approx. 10–14) with no prior exposure to Mandombe. After consent procedures, learners were randomly assigned by classroom to one of two conditions:

Mandombe group (M): 25 learners

Latin group (L): 25 learners

Randomisation was at class level to respect school schedules; within each school, we matched classes by grade and general academic level as reported by the staff. But teachers did not have access to the questions (handed to students in envelopes and proctored by our skeptical reviewer as part of ATSS standards on separation of duties).

3.2 Design

The study consisted of a single 3-hour module during the school day, structured as:

1. Shared ethics mini-lesson (≈ 30 minutes)

2. Script introduction and practice (≈ 90 minutes)

3. Written task and short debrief (≈ 30–40 minutes)

Both groups received the same ethical content in part 1, framed in simple language around cooperation, promises, support and fairness. In part 2, they differed only in the script they practiced. Part 3 used identical written questions, with only the visuals changed.

3.3 Script introduction

3.3.1 Mandombe group

The M group learned:

the ambiguous mvuala-shape;

the kisimba corresponding to that mvuala;

how attaching the kisimba resolves which consonant it is;

how the kisimba’s orientation follows the mvuala;

one simple combination (e.g. a mvuala mpamba).

The teacher, following a short prepared script, explained that:

the mvuala-shape alone “does not yet show clearly who it is”;

with the kisimba “it becomes itself”;

mvuala and kisimba “depend on each other”;

combining signs creates a stronger, more stable unit.

Learners spent most of the 90 minutes copying mvuala, attaching kisimba, and forming simple combinations, with individual correction on shape and orientation.

3.3.2 Latin group

The L group learned:

the letter “b” in the standard school script;

a second version of “b” drawn inside a simple square box.

The teacher used the same ethics story but now linked it verbally to the idea that “letters together make words like people working together in a team.” The letter “b” and its boxed variant were presented neutrally as “the same letter in a different situation.” No meaning was attached to the box beyond “around” or “inside.”

Learners practiced writing “b” and “b in a box” and stringing letters into short syllables or pseudo-words.

3.4 Written task

At the end of the module, each learner received a one-page worksheet. At the top were two printed signs:

For M group:

left: mvuala-shape alone;

right: mvuala with its kisimba in canonical orientation.

For L group:

left: “b”;

right: “b” drawn inside a square box.

Below the signs, the following three questions (in French) appeared, and were read aloud:

1. Q1. What is the difference between the two signs?

(One or two sentences.)

2. Q2. Imagine each sign is like a person. What kind of person could the first sign be, and what kind of person could the second sign be?

(One or two sentences.)

3. Q3. Is one of these people in a better situation than the other? Why?

(One or two sentences.)

Learners wrote their answers individually. Teachers were unable to suggest ideas (absent from the mini test. Completed sheets were collected immediately by the research team.

3.5 Coding

Our primary interest was whether learners used the structure of the script to think about relations and responsibility. We therefore defined two simple coding axes, applied to each learner’s full set of answers.

3.5.1 Axis 1: Epistemic stance

E0 – Literal / flat:

The learner treats the sign as fully determined and does not mention uncertainty, lack of information or “missing something.” Examples:

“Here it is b, here it is b in a square.”

“It is a person, any person, a child or a grown-up.”

E1 – Ambiguity-aware / questioning:

The learner explicitly notes that the sign alone is unclear, incomplete or “not yet itself,” or questions the situation. Examples:

“I cannot tell what this mvuala is by itself.”

“This one is alone, missing something.”

“Why is this mvuala alone?”

3.5.2 Axis 2: Type of person mapping

P0 – Trivial / generic person:

The learner names a type of person with no social, relational or ethical content. Examples:

“It can be a child or an adult.”

“It is a person, any person.”

P1 – Boxed / stereotyped mapping:

The learner uses the box or fixed shape to put a person into a rigid category or container, without exploring relations. Examples:

“This is a child in a room / in school / in a car / in a coffin.”

“This one is just a b, always a b.”

P2 – Relational / ethical mapping:

The learner uses alone/with, missing/complete, or inside/outside to speak about relations, support, belonging, exclusion or responsibility. Examples:

“The mvuala alone is like a person without friends or family.”

“I do not know what it is because it has no kisimba; that is like someone who did not get education or was always alone.”

“With the kisimba it is balanced, like a person who finds their group and can help and be helped.”

Two coders coded all 50 sheets independently while blind to condition (sheets were shuffled and group labels removed). Disagreements were discussed until consensus. For the main contrast of interest (presence vs absence of relational mapping P2), agreement exceeded 85%.

3.6 Analysis

The primary quantitative comparison was the proportion of learners in each group who produced a relational mapping (P2). We also examined the distribution of epistemic stance (E0 vs E1) descriptively. Given the sample size and exploratory nature, we used simple proportion tests and focused on effect sizes and patterns rather than fine-grained statistics.

4. Results

4.1 Relational mappings grounded in script structure

In the Mandombe group, a clear majority of learners produced relational / ethical mappings that explicitly used the mvuala–kisimba structure (P2). For example:  “This mvuala alone is unclear, I cannot tell who it is. It is like a person without family or friends. When we put the kisimba, it becomes itself; that is like someone who finds their group and can help others.”  “Alone it is naked and imbalanced, like a child who is always alone. With the kisimba it is dressed and stable, like someone who is in a family.”

By contrast, in the Latin group, most learners either gave generic person labels (P0) or simple container descriptions (P1) for the boxed “b”:

“This is just b and here it is b in a box. The person is a child in a room.”

“It is a person in a car or in a coffin. It is still a b.”

Very few Latin learners used the difference between the two signs to speak about support, belonging or responsibility.

Numerically, the pattern looked as follows (no claim of generalizability is made at this stage. Treat as evidence of existence, as we do not measure prevalence):

M group: 18 of 25 learners (72%) coded as P2, 4 as P1, 3 as P0.

L group: 5 of 25 learners (20%) coded as P2, 9 as P1, 11 as P0.

A simple comparison of proportions suggests a large difference in the likelihood of producing relational mappings grounded in script structure. Even with conservative assumptions, the effect size is consistent with our hypothesis H₂.

4.2 Epistemic stance: noticing ambiguity

On the epistemic axis, many Mandombe learners spontaneously highlighted the ambiguity of the lone mvuala and the clarifying role of the kisimba (E1). Examples include:

“The first one is alone so I do not know exactly what mvuala it is.”

 “Without the kisimba I cannot be sure, it is like someone without explanation.”

In the Latin group, E1-type statements were rare. The most frequent pattern was E0:

 “The first is b, the second is b with a square.”

 “It is just b, it stays b.”

A small number of Latin learners extended the box metaphor to “being stuck” or “locked in,” but usually without questioning identity or knowledge (“it is in a coffin,” “in a room”). For this initial study, we did not quantify E0/E1 with the same rigour as P2; instead we note the contrast qualitatively and treat it as an emergent pattern.

4.3 Summary

On our primary outcome, the data support a cautious but clear conclusion: after a brief module with identical ethical content, learners in the Mandombe condition were substantially more likely to use the script’s internal relational structure (mvuala alone vs mvuala + kisimba) to think and talk about relations and responsibility than learners taught with a Latin-based script. This matches informal classroom observations from longer Mandombe courses and provides the first controlled classroom evidence for Mandombe’s ethical affordance at the level of symbolic cognition.

5. Discussion

5.1 What we can reliably claim

Our strongest claim is structural. Mandombe’s mvuala–kisimba system encodes identity as a relational resolution of ambiguity. A base shape does not fully “say who it is” until it enters into a mutual orientation with a kisimba and with other glyphs. Latin grapheme structure does not exhibit this kind of intrinsic ambiguity–resolution.

At the classroom level, this study shows that novice learners can and do mobilise this relational backbone almost immediately when invited to think about persons and responsibility. After only three hours, many Mandombe learners treated the lone mvuala as underdetermined and used the addition of the kisimba as a metaphor for support, belonging and shared responsibility. Latin learners, working with an equally short introduction to “b” and “b in a box” and hearing the same ethical story, rarely did so. Their answers remained largely literal, generic or boxed.

We therefore maintain a modest but important conclusion: Mandombe provides a relational geometric backbone which, when combined with its traditional philosophy, supports an ethical learning environment that a Latin-based script is structurally poorly equipped to reproduce. We make no claim that this immediately alters behaviour or clinical profiles. We do claim that it changes what kinds of ethical metaphors are natural and available, even for novices.

5.2 Educational implications

For African education systems that remain structurally dependent on colonial scripts and curricula, this finding has practical implications. Ethics and civic education are often taught as separate “content,” added in a few hours per week, while the core symbolic infrastructure of schooling (script, language, grading scales) continues to rehearse individualised, competitive and rigid categories.

Mandombe offers a built-in different option. Because its geometry and its philosophy are aligned, it allows teachers to integrate ethical discussion into the basic act of writing. A child does not only learn that “this shape is this sound.” The child learns that identity appears in relation, that “alone” is unclear and unstable, and that mutual orientation matters. Our data suggest that even a short module can activate this logic. For longer, full pathways of Mandombe education, the cumulative effect on ethical perception and social imagination is likely to be stronger and should be studied.

5.3 Relation to DSM-H and structural pathologies

DSM-H and related frameworks argue that antisocial disorders in African contexts are often shaped by structural and symbolic environments rather than only by individual pathology. Systems that reward domination, rigid categorisation and dehumanising labels contribute to what has been called the colonised mind and the colonised Dark Tetrad.

Within this perspective, scripts and symbolic systems are not neutral. They either support, or silently undermine, relational personhood. Our results do not yet speak to clinical outcomes, but they show that Mandombe can be used to rehearse relational identity and shared responsibility at the symbolic level. This makes it a plausible candidate as one “upstream” protective factor in a multi-layered DSM-H strategy. Formal tests of such protective effects will require larger, longitudinal and multi-method studies.

5.4 Emergent ambiguity and stereotyping: an agenda for separate study

During coding we noticed a second pattern, which we did not design the study to test. Mandombe learners repeatedly described the lone mvuala as “not yet itself,” “missing something,” “unclear,” and used the kisimba as a way to resolve that ambiguity. Latin learners very often took the identity of the “b” as self-evident and treated the box as a container that places any person in a room, a car or a coffin, without questioning identity or the adequacy of the label.

This suggests a possible difference in epistemic stance (awareness of ambiguity vs flat certainty) and in the use of shapes as labels or boxes. The Latin group sometimes produced answers that resembled early forms of stereotyping: a clear-cut box with a fixed label, regardless of context. The Mandombe group often resisted premature closure: “I cannot tell what it is by itself; it needs something to be clear.”

We judge this pattern intriguing but complex. It touches on deep questions about how children learn to navigate uncertainty, how they adopt or resist rigid labels, and how scripts may interact with colonised or decolonised forms of categorisation. Given the potential implications, we prefer not to treat it as a result of this study. Instead, we flag it explicitly as a hypothesis for a separate, dedicated project, with its own design, coding scheme and safeguards.

5.5 Limitations

Several limitations must be acknowledged.

1. Short intervention and small sample. Our module was only three hours, with 25 learners per group. This is suitable for detecting large, mechanistic differences but not for subtle effects or individual trajectories.

2. Randomisation by class, not individual. We randomised at classroom level for logistical reasons. Teacher and classroom effects cannot be fully ruled out, although both groups followed the same ethical script with close monitoring.

3. Written responses only. We captured language about ethics, not behaviour. We do not know whether children who use relational metaphors in writing act more responsibly in practice.

4. Single contrast script. We compared Mandombe to a Latin-based script. Other scripts with rich geometric structure might show different patterns and deserve attention.

These limitations suggest that our findings should be read as an initial, well-controlled demonstration of a symbolic affordance, not as a comprehensive evaluation of Mandombe’s educational or clinical impact.

5.6 Future research

Three directions follow naturally.

1. Larger and replicated classroom studies. Replicate the design in multiple schools, with more classes and possibly variant tasks, to confirm the robustness of the relational mapping effect and test how much ethical discourse is needed to activate it.

2. Multi-method designs. Combine symbolic tasks with simple behavioural measures (e.g. resource-sharing games) and teacher reports to see whether children who use Mandombe’s relational metaphors also show more cooperative or fair behaviour.

3. Longitudinal and DSM-H–aligned work. Follow cohorts of learners in Mandombe and non-Mandombe schools over several years, combining symbolic tasks, psychometric measures aligned with DSM-H constructs (e.g. colonised-mind indices, antisocial traits) and qualitative data. This would allow us to test whether Mandombe’s relational backbone functions as a protective symbolic environment or whether its effects remain local to classroom discourse.

A fourth line of research, motivated by the emergent finding, would focus specifically on ambiguity, labelling and stereotyping across scripts. Here the outcome would not be “ethical metaphors” but patterns of how children handle uncertainty and assign labels, a question that reaches into the core of colonised cognition.

6. Conclusion

This study sits at the lowest tier of a broader ATSS1.2 agenda. We do not claim that Mandombe cures antisocial disorders or directly reforms institutions. We show something simpler and more basic: that Mandombe’s mvuala–kisimba geometry, when taught together with its traditional philosophy, gives learners a concrete visual and symbolic language for thinking about relations and responsibility that a Latin-based script does not easily provide.

In a short, controlled module with identical ethical content, novice learners in the Mandombe condition were much more likely to describe the lone mvuala as “not yet itself,” “missing something,” and to use the addition of the kisimba as a metaphor for support, belonging and shared responsibility. Latin learners, using “b” and “b in a box,” mostly stayed at the level of literal description or generic, boxed labels.

Even at this modest level, the implications are significant. In societies marked by structural antisocial pathologies and colonised symbolic environments, the choice of script is not neutral. A script can silently rehearse isolation and fixed labels, or it can rehearse relational identity and mutual orientation. Mandombe belongs to the second category. It does not merely write words. It draws, in its very lines and rotations, an ethic of being a person with and through others.

References:

Foundational:

Wabeladio P. (1984) Méthodologie de l’enseignement du Mandombe;

Nsiangani K. (2025) The Colonized Mind.

Nsiangani K, Corpus fondateur tu MEN-D, 2021, 2025 editions, USK Kinshasa

Theoretical

[1] Mbiti, J. S. (1990). African religions & philosophy (2nd rev. & enl. ed.). London: Heinemann.

[2] Menkiti, I. A. (1984). Person and community in African traditional thought. In R. A. Wright (Ed.), African philosophy: An introduction (3rd ed., pp. 171–181). Lanham, MD: University Press of America.

[3] Gyekye, K. (1995). An essay on African philosophical thought: The Akan conceptual scheme (Rev. ed.). Philadelphia, PA: Temple University Press.

[4] Pae, H. K., & Beckett, G. H. (2024). Script relativity in psycholinguistics and sociolinguistics: Writing systems as a catalyst for cognition and culture. Docens Series in Education, 6, 28–47.

[5] Pae, H. K. (2020). Script effects as the hidden drive of the mind, cognition, and culture. New York, NY: Springer.

[6] Winskel, H. (2022). Script relativity hypothesis: Evidence from reading with different spatial layouts and varied lexical tones. Reading and Writing, 35(6), 1323–1341.

[7] Nsiangani, K. (2014). The Dark Tetrad Traits of Empire. [Monograph].

[8] Nsiangani, K. (2016). From Mvemba Nzinga to Modern Puppets: Cultural Violence and Colonial Leadership Pathologies. [CEMA, USK, Monograph].

[9] Nsiangani, K. (2025). DSM-H: Diagnostic and Statistical Manual of Human Structural Pathologies. [CENA Editions, Oct 2025 monograph].

[10] Nsiangani, K. (2025). The Colonised Mind: A Clinical Framework. [1st sem Lecture MEN-D dpt].