I studied Holland, Holyoak, Nisbett and Thagard's 1986 Induction: Processes of Inference, Learning, and Discovery in close detail not long after it came out, but I have not had occasion to return to Holyoak's work on analogy since. That's unfortunate (and in a different more philosophical life I would most definitely have kept current on this literature) because it renders even more disappointing this paper which seeks to explain the role of analogy in the teaching of internet (and) information literacy. Holyoak approaches analogy from a psychological perspective, which appeals to me, as a literal or logical understanding oversimplifies some of the nuances. The sort of explanation Ferranti is attempting here is much more successfully presented in the first few paragraphs of Holyoak here. Holyoak and Gick's use of schemas to support analogy, referenced by Ferranti, is found here. Just scanning his work suggests to me that Holyoak has come a long way from these roots - look at A Symbolic-Connectionist Theory of Relational Inference and Generalization, for example (definitely read the concluding section on 'The Power of Self-Supervised Learning'). It's a pity the Ferranti paper is so poor, given the rich tradition it is drawing from. Part of that is created by a difficulty of the language - the paper is very awkwardly written. But part of it is also created by a representation of a complex and interesting phenomenon such as anlogy with the simplistic sort of schema (in this case Gentner's4 levels of similarity) all too common in writings on education.
Today: 10 Total: 103 [Direct link] [Share]
