A field does not become stronger by surrounding itself only with compatible ideas. Agreement can consolidate vocabulary, but confrontation tests boundaries. It forces a concept to specify what it can explain, where it fails and which neighbouring account it must resist in order to remain meaningful. Intellectual confrontation is therefore different from polemic. Its objective is not victory but precision. Lefebvre and Foucault converge around the production of reality, yet distribute explanatory weight differently between social practice, discourse and institutional formation. Habraken’s differentiation of levels can clarify complexities that network approaches risk flattening, while attention to mediation can challenge architectural models that assign control too neatly. Benjamin’s montage resists the systematic closure toward which any conceptual architecture may drift. Bowker and Star expose the exclusions that even well-designed categories produce. These tensions are productive because no position becomes the universal solvent of the field. Confrontation also protects against dependency. A field that only borrows from authorities remains subordinate to their questions; a field that confronts them can inherit operations without accepting their limits as final. This requires fairness: a position must be reconstructed in its strongest form before it is criticized. Concepts should be tested against their capacity to clarify mechanisms, preserve complexity, support evidence and generate consequences. Confrontation is structural stress testing. What survives becomes more precise; what fails reveals the boundary of the problem. A mature field does not conceal these failures. It uses them to distinguish its own grammar from the languages through which it first learned to speak.