The inverted genotype and its implications for the flexibility of architectural models

Sean Hanna


Most current methods of architectural representation, in the form of CAD and its derivatives, developed during a period of intense activity in both molecular biology and artificial intelligence, with which they shared a fundamental principle in the emphasis on a symbolic code as a generating and transmission mechanism. This basis persists, and is further strengthened, in contemporary developments of parametric modelling, design scripting and building information modelling. While this yields powerful tools, it also carries inherent disadvantages in terms of potential insensitivity to outcomes and inflexibility beyond set constraints—when design activity is focused on the code, it may be more difficult to take account of the result. It is apparent that Space Syntax has contributed a means for analysis to mediate this, but this paper will highlight Hillier and Hanson’s crucial notion of the ‘inverted genotype’, and most importantly the proposal of an information ‘retrieval mechanism’, in offering an alternative to the primacy of the representation. This novel contribution was that the information carrying element is not a coding mechanism held in and transferred via a description centre (the genotype), but in the space itself (the phenotype). In its proposition of a retrieval mechanism to read this information, this runs counter to the ‘central dogma’ of molecular biology, which allows information flow only from genotype to phenotype. In its proposition that the information is transferred not by the code but by its embodied result, this runs counter to what was, at the time, the dominant paradigm in artificial intelligence and the philosophy of mind, and is still the basis for our CAD and related architectural representations today. The computational implementation of such extraction mechanism is more difficult, but the basis of this is demonstrated with reference to a number of current and recent models. This paper will make the case for such models in allowing communication without standards and permitting creative change.


Genotype, Representation, CAD, Parametric Model, Creativity

Full Text: JOSS_2011_P247-272.pdf