Page 65 - JCAU-6-1

Journal of Chinese
Architecture and Urbanism Exploring abduction in regenerative design

a value. The upper bound on the design outcome is already abduction is a quasi-causal process inferring from the
fundamentally determined by the original database. consequence (usually a surprising fact) to its precondition,
However, as an innovation rather than an improvement, inquiring what entails this effect. Abductive reasoning
there are no pre-existing and comparable references for the does not point directly to an absolute conclusion but tends
DeepGreen protocols. to find explanations. Thus, the abductive conclusion has

The GAN_Physarum model in the DeepGreen project some residual uncertainty and is still a hypothesis to be
draws form-driven inspiration from slime molds and verified (Sober, 2019).
trains an AI to project the network onto urban forms Design reasoning inherits reasoning models from
(Figure 7). Elements that do not originally exist in the scientific reasoning. Commonly, deductive reasoning
urban typology are absorbed and transformed into the new arrives at design outcomes from explicit and approved
urban morphology, altering the initial design expectations. premises to show what must be. Inductive reasoning
This process of ideation is not delimited by pre-existing generalizes what literally is by summarizing specific
paradigms and experiences, and thus, its logic differs from examples. Abductive reasoning offers new recipes and
deduction and induction. It is clear that these two models redefines the desired value in design, suggesting what may
of reasoning are inadequate to explain how new knowledge be (Dong et al. 2012). However, in Roozenburg’s opinion,
and ideas arise. This is why Charles Sanders Peirce Peirce specified two abductive modes. He re-articulated
developed abduction as an extended and comprehensive the two patterns as explanatory abduction and innovative
reasoning mode to create new hypotheses and knowledge abduction. The first pattern is interpreted as the following
(Burks, 1946). Peirce exemplifies the variations among syllogism (Roozenburg, 1993):
the three reasoning modes as presented in Table 1 (Peirce,
1878): Premise q a given fact, a desired result
“Deduction is the inference of a result from a rule and a Premise p → q a given rule, IF p THEN q
case (...) Induction is the inference of a rule from a case and
a result (...) Abduction is an inference of a case from a rule Conclusion p the cause (I)
and a result” (Roozenburg, 1993, p. 9). In fact, common
modes of reasoning infer from a cause to its effect, while In explanatory abduction, the rule is known and
becomes a premise, allowing us to infer from the result to
the cause. Roozenburg commented that the abduction in
Syllogism I “is not about discovery but about diagnosis or
troubleshooting” (Roozenburg, 1993, p. 10). In contrast,
innovative abduction involves the conception and inference
of the rule along with the cause without assuming the rule’s
truth. This pattern is, therefore, expressed as (Roozenburg,
1993; Kroll & Koskela, 2017):
Premise q a given fact, a desired result

Conclusion p → q a rule to be inferred, IF p THEN q

Conclusion p the cause (II)
Since the hypothesis enabling us to infer from result
to cause (p→q) remains to be supplemented, “the essence
Figure 7. GAN_Physarum: Paris (scale: 1 km). Source: ecoLogicStudio, of abduction lines in finding or forming the missing
GAN_Physarum: la dérive numérique, 2022 hypothesis” (Roozenburg, 1993, p. 9). Peirce, in particular,

Table 1. The variations among the three reasoning modes exemplified by Peirce

Reasoning mode Rule Case Result
Deduction All the beans from this bag are white. These beans are from this bag. These beans are white.
Induction These beans are from this bag. These beans are white. All the beans from this bag are white.
Abduction All the beans from this bag are white. These beans are white. These beans are from this bag.

Volume 6 Issue 1 (2024) 7 https://doi.org/10.36922/jcau.1084