This changes the meaning of comparison. Comparison cannot serve only to place cities along a common scale of development, sustainability, governance or technological maturity. Robinson’s comparative practice turns the relation among cities into a procedure for destabilising the categories used to know them. Watson’s clash of rationalities performs the same operation inside planning: expert systems and lived urban economies do not merely disagree about solutions; they organise reality through different forms of reason. Pieterse intensifies this plurality by holding technical policy analysis, political economy and ethnographic urbanism in epistemological tension. The city is irreducible because no single register exhausts its organisation. Irreducibility does not mean incomparability. Simwanda’s four-city heat analysis demonstrates precisely the opposite. Shared measurements make relations visible, but their value depends on retaining the distinct configuration of Lagos, Nairobi, Addis Ababa and Lusaka. A common sensing apparatus reveals divergent combinations of impervious surface, green space, morphology and temperature. Comparison works when it identifies a mechanism without converting one manifestation of that mechanism into the norm. The comparative object is therefore neither the isolated city nor the universal model but the relation between common processes and situated configurations. Situated knowledge becomes infrastructural when institutions are built to sustain that relation. Cape Town’s CityLabs show that co-production requires more than consultation. It demands organisational spaces where disciplinary expertise, municipal practice and broader social knowledge can alter one another. Okyere’s work in Accra performs a similar operation through participatory mapping and interviews: the walking environment is reconstructed from the accounts of those whose mobility is statistically common yet politically marginal. This is epistemic redistribution in practical form. The resident does not merely validate a technical diagnosis; lived experience helps constitute the diagnosis.
The African city then ceases to function as a corrective footnote to urban theory and becomes a generator of methodological demands. It requires comparative practices capable of starting from anywhere, planning theories able to recognise multiple rationalities, data systems that do not confuse absence of official records with absence of knowledge, and environmental models that preserve morphological and climatic difference. It also requires attention to the infrastructures through which knowledge travels: universities, city governments, participatory platforms, remote-sensing systems, workshops, policy documents and digital networks. Each grants visibility to some urban relations while suppressing others. Socioplastics can treat this as an archival principle of non-homogenisation. Sources from African cities should not be indexed merely under a regional label, because “Africa” is not the explanatory variable. Each object should preserve city, method, positionality, scale, knowledge producer, infrastructure, climatic condition, mobility regime and institutional context as active relations. Comparison can then occur across the graph without destroying local specificity. Accra can speak to Madrid through walkability without becoming equivalent to Madrid; Lagos can enter a heat comparison without becoming representative of “African heat”; Cape Town can provide a model of knowledge brokerage without being elevated into a continental template. The archive becomes comparative precisely because its nodes remain irreducible. The strongest consequence is epistemological humility made operational. A robust global urban field is not one that finally discovers the universal grammar behind all cities, but one that builds procedures through which concepts remain revisable by the places they encounter. African urbanism contributes here not as a bounded regional specialism but as a powerful demonstration that theory changes when authority over abstraction is redistributed. The knowledge graph becomes stronger when difference can modify structure rather than merely occupy it; within Socioplastics, this constellation condenses as RegimeMutation.
