Friday 9 August 2019

John Bateman On All Models Being Wrong By Definition

…fuzziness in the model[l]ing means that fuzziness in the modelled may become inaccessible because multiplying fuzziness by fuzziness is pretty much the same as multiplying not-fuzziness by fuzziness. better to have a non-fuzzy model that is wrong (easily done, by definition, because all models are wrong by definition) that can show real fuzziness rather than imagined fuzziness. difficulty may always be interesting; fuzziness often not. …

Blogger Comments:

[1] Any reader would be very hard-pushed to find a definition of a scientific model that includes the interpersonal assessment that all models are "wrong".  In science, models are evaluated as valid or not, according to whether, for example, they are self-consistent or not, consistent with data or not, explanatory or not, predictive or not, and so on.

[2] On the other hand, fuzziness (indeterminacy) in language is of great interest to Halliday & Matthiessen (1999: 547-62), who, as well as identifying 5 types of indeterminacy, explain its significance. Halliday & Matthiessen (1999: 547-8, 549):
What does it mean to say that a natural language is an indeterminate system? In the most general terms, it suggests that the generalised categories that constitute language as a system — as "order", rather than as randomness or "chaos" (let us say randomness rather than chaos, since chaos in its technical reading is also a form of order) — are typically not categorical: that is, they do not display determinate boundaries, fixed criteria of membership, or stable relationships from one stratum to another. We could refer to them as "fuzzy", in the sense in which this term is used in fuzzy logic, fuzzy computing, etc.; but we prefer to retain the term "indeterminate" for the phenomena themselves, since "fuzzy" is usually applied to the theoretical modelling of the phenomena (it refers to meta-fuzz rather than fuzz). …  
We have tried to make the point that the human condition is such that no singulary, determinate construction of experience would enable us to survive. We have to be able to see things in indeterminate ways: now this, now that, partly one thing, partly the other — the transitivity system is a paradigm example, and that lies at the core of the experiential component of grammar.

No comments: