Imagining the imaginable: a reinterpretation of the function of economists' concern about structural isomorphism in economic theorizing

Abstract

By using a metatheoretical interpretation of the development of international trade theory as an example, I illustrate that, as is manifested in the practices of economic theorization, a theoretical representation can be decomposed into two component representations: a formal representation and a causal narrative representation. I further maintain that, with respect to both component representations, the concern of isomorphism is important in that it is the guiding idea that underlies economists' practice of identifying both an adequate formal model and a plausible causal story to represent the targeted phenomenon. As a result, it can be argued that the nature of the representational relation is manifested in the practices of economists as they imagine the imaginable causal stories on the basis of imaginative causal structures depicted in the causal models of the targeted phenomenon.

Keywords:

• economic theorizing
• representation
• semantic approach
• structural isomorphism
• causal narrative
• imagination
• causal structure
• international trade theory

Acknowledgements

I would like to express my gratitude to those who have helped me shape the present form of this paper. The main argument of this paper was completed during my research leave between February and September 2009. During this time, I first stayed at the Institute of European and American Studies at Academia Sinica (IEASAS), Taiwan, for six months' research. I express my appreciation to my sponsor – Professor Wan-Chuan Fang – and Professor Norman Y. Teng and Dr Jih-Ching Ho for their hospitality and for helping me quickly adjust to their outstanding research environment. Then I moved to the Center for Humanities Research at National Science Council, Taiwan, for another two months' research and, at the end of the stay, finished the first draft of this paper. I thank my mentor – Professor Cheng-Hung Lin – and Mr Wei-Hsiung Huang for their always helpful assistance during my stay there. To all of my colleagues who shared my teaching load during my absence, I would like to say that their support made my research leave possible, and I want to express my great appreciation to them. Special thanks go to my colleague Dr Chi-Chun Chiu, who helped me clarify some logical intricacies in the previous version of the paper so as to make the revised paper more readable and convincing. Also, I would like to thank my colleague from the Economics Department, Dr Hsiang-Ke Chao, for his long-term academic conversation and advice. The completed draft of this paper was presented at the Workshop on Body and Meaning IV held at IEASAS in November 2009 and received helpful comments from Professor Ruey-Lin Chen, and I should also express my thanks to him. Finally, I have benefited much from the encouragement and critical comments offered by two anonymous referees; their suggestions and guidance have helped to greatly improve the previous argumentative structure of the paper, and so they too have my great appreciation. This paper is the result of a research project supported by National Science Council, Taiwan, project number 97-2410-H-007-042.

Notes

 1. According to Rol, when theorists conduct idealization, they assign limit values to the causally relevant variables of an equation or function. Take our equation q₁ = 8.5–0.85p₁–0.12p₂+0.63p₃ as an example. There are at least three different ways that limit values can be introduced into the equation: (1) to add a ceteris paribus clause: in this equation, p₂ and p₃ are ‘other circumstances’ (ceteris); by adding this clause, it makes Δp₂ = 0 and Δp₃ = 0; (2) to add a ceteris absentibus clause: it will make p₂ = 0 and p₃ = 0; (3) to add a ceteris neglectis clause: it will make these two partial differentials ∂q₁/∂p₂ ≈ 0 and ∂q₁/∂p₃ ≈ 0, meaning that the causal influence of the ceteris, p₂ and p₃, with respect to q₁ is all negligible. For more details, refer to Rol (2008, pp. 70–71).

 2. For more details about Kripke-Lewis semantics, refer to Kripke (1980) and Lewis (1973).

 3. Leontief's investigation showed that US import-competing production required 30% more capital per worker than US export production; this conclusion was just the opposite of what the H-O model predicted.

 4. These issues include the topic of the change of the price of production-factor among trading nations, such as the condition of factor-price equalization; and that of the change of the living standard of trading nations, such as the change of the living standard of a nation as a whole and its suppliers of the factors that are used intensively by the import-competing industry once the protection policy – e.g., a high tariff levied on imported goods – is launched.

 5. With respect to the strategy of the weakening of conditions and its logical effect, Hamminga writes, ‘The weakening of conditions result can be “plausibility”-interpreted in an analogous way: Suppose in V_lmn, IT can be proven from C₁, …, C_k. If we now find weaker conditions for IT in V_lmn, the class of worlds in which IT holds must because of that have been widened. Weaker conditions always permit more cases, i.e. more worlds in which the conditions are satisfied. Again, IT holds in more worlds than we previously knew, and thereby has become more plausible’ (1983, p. 73).

 6. Emphasis added.

 7. The quotation in the text is excerpted from Hamminga (1983). For the origin of the quotation, please refer to Stolper and Samuelson (1941).

 8. The quotation is also from Hamminga (1983); for the origin of the quotation, please also refer to Stolper and Samuelson (1941).

 9. Hamminga regards FEA-conditions and EI-conditions as being unchangeable or irremovable. He supposes that when economists are conducting model-manipulation, they manipulate only their theory's special conditions and leave intact the FEA and EI-conditions. Whether to follow this convention does not affect the validity of my argument, so for the sake of consistency in expressing my own idea, I do not follow the convention but allow that all the conditions, including FEA, EI, and special conditions, are changeable or removable.

10. Although Hamminga does not state this idea in these words, we can attribute the idea to him. Recall Hamminga's words in section 4: ‘“the plausibility of a theorem” can be taken to mean “the probability of the theorem being true in the real world”, where the “real world” is just one of the worlds that can be expressed by means of the economists’ language (production functions, utility functions, factor endowments)' (1983, p. 71). If ‘the worlds’ in the quotation are interpreted as ‘many putative worlds’ that are constituted by various conditions-combinations – e.g., by various production functions and utility functions with different properties and shapes – then ‘the set of the worlds’ can be regarded as ‘the huge hypothetical world’ mentioned here; the huge hypothetical world is a huge composite formal structure consisting of a class of component models, and each component model represents a particular putative world constituted by a specific set of conditions-combination. We can thus conjecture that Hamminga would endorse the idea that ‘the real world – i.e., the actual world – can be any one of the various models – i.e., any one set of the conditions-combinations’ as mentioned in the text. This interpretation thus justifies my attribution of the idea to Hamminga.

11. Emphasis added.

12. Emphasis added.

13. Emphasis added.

14. Emphasis added.

15. Emphasis added.