Navigation – Plan du site
Revue des livres
Essais critiques

A Reply to Paul Dumouchel

Maurice Lagueux
p. 247-251

Texte intégral

1I am grateful to Paul Dumouchel for his very thoughtful presentation of many aspects of my work and for the important questions he raises about other ones that definitely require clarifications. What must be clarified bears mostly on the relations between natural selection and rational explanations, but also on the scope and future of the latter.

2First, let me point out that I do not claim that natural selection and rational explanations are as radically exclusive as Dumouchel suggests. I even noted explicitly that “natural selection does not exclude motivated behaviour” (181). The only thing that I excluded in the rest of the sentence, which concerns its eventual application to economics, was “reliance on wise decisions when it comes to explaining success or survival.” Yet, I must admit that my wording was possibly a source of confusion, especially when I added: “at some point, economists must choose between explanations based on natural selection and those based on rationality” (and possibly also in the too schematic table on p. 195). However, I purposely postponed further precisions about this sentence to the end of the chapter where I explain that my point was not that natural selection excludes rationality and conversely, in the sense that one should globally opt either for one or the other. My point was rather that if a particular phenomenon—such as a remarkable equilibrium resulting from unintentional interventions—is fully explained by natural selection, any attempt to explain the same phenomenon by rationality would be senseless. Conversely, I added that if a particular phenomenon, such as “a free economy, animated by entrepreneurs who make self-interested decisions, generates markets relatively well equilibrated in a way that is judged far superior (measured by the precision and the rapidity of the adaptation required) to anything that natural selection might eventually generate(197) then rationality should be called in to explain the phenomenon, given that any attempt to explain it by natural selection would be unsuccessful by hypothesis. I suggested that explaining the same phenomenon with both kinds of explanation would result in an unintelligible overdetermination and I attempted to illustrate this with a few examples drawn from economics and other domains (196-199).

3Having little propensity for uttering anathemas, I would never claim that Darwinian evolutionary explanation should be “ruled out a priori from economics.” Actually, on pages 196-197, I did not argue, as Dumouchel claims, that Darwinian explanations “could only be applied to economic phenomena if entrepreneurs acted ‘perfectly randomly’.” What I rather said is that if the phenomenon to be explained “would result from totally irrational actions from entrepreneurs … , then only a natural selection mechanism … might do the job.” This formulation does not preclude this kind of explanation from being applied in other circumstances as well, but it is true that I have not provided examples of such situations. Dumouchel offers such an example by interpreting Nelson and Winter’s analysis as “the description of a typically Darwinian process.” For my part, I referred rather sympathetically to the way Nelson and Winter emphasised the role of routines in enterprises and I agreed (183) with their own insistence on the idea according to which their view was “unabashedly Lamarckian,” given their “espousal to Lamarckianism” (Nelson and Winter, 1982 ,11). I do not exclude that Dumouchel might be right against Nelson and Winter when he suggests that their approach is “typically Darwinian,” but I think that this is a matter of interpretation. Dumouchel acknowledges that those routines might be chosen for various reasons, but for reasons usually related to local preoccupations, which “generally have little to do with the overall goal of maximizing profit or market share.” It is difficult to appreciate to what degree these reasons are more or less directly related to this overall goal, but if Dumouchel is right in claiming that “relative to the goal of the enterprise’s success they constitute random variation which are produced by a relatively blind mechanism,” one should conclude, in conformity with what I said about explaining such a particular phenomenon, that this Darwinian explanation is not compatible with an explanation based on rationality. Moreover, were Nelson and Winter’s Lamarckian interpretation correct, it would imply that a rational explanation would no longer be excluded. Indeed, as I have suggested (195), a Lamarckian approach to economics is not that far from the classical one promoted by Adam Smith who, incidentally, was very aware of the important role of local institutional factors similar to those put forward by these authors.

4According to Smith and his followers, capitalists are just rational enough to decide to increase or reduce production depending on the importance of their gains or losses, and also just rational enough to transfer their capital in a more rewarding industry when the prospect are too dim in their own industry; this explanation also supposes that workers and landowners behave in a similarly rational manner. My point is not to offer this as a model for any economic explanation but to draw attention to the fact that, even in such explanations and in their modern reformulations, these rational agents, who are engaged in fierce competition among themselves, cannot control the outcome of their decisions, the consequences of which being largely unwanted. In some sense, one might say that the relatively efficient market that results from the complex interaction of all these rational decisions is also “produced by a relatively blind mechanism.” In any case, one cannot explain the phenomenon by saying both that the result being blindly attained, natural selection is required to explain it and that, the mechanism being not so blind after all, a rational explanation can work. In the tradition derived from Smith, it was understood that consequences are largely unwanted because agents do not look for the efficiency of the market, but are not attained blindly to the point of excluding an objective relation between their rational decisions and the result attained. This is the reason why it would be odd to try to explain this particular phenomenon with natural selection.

5Clearly, every selection is not an instance of natural selection as illustrated in my book by the example of the selection committee in which both applicants and committee members make intentional decisions (187-188). Dumouchel’s example of the class of children who already know how to read is similar and even better than mine. True, it illustrates that selective explanations target the whole population rather than individuals, but also that the selection involved in such a case is not natural selection, since it supposes intentional decisions both from children or their parents and from the kindergarten’ administrators who decided to restrict admission. On this ground, the Smithian explanation of the market’s efficiency at providing goods required by a society goes further. It implies that this blindly generated efficiency, resulting from the competition between capitalists who look for profits and not for any specific feature of the market, does not result from natural selection either. I can hardly imagine how claiming the contrary would not be dissolving the meaning of natural selection uselessly.

6The explanatory power of natural selection rests on the fact that an exceptionally remarkable order was realised without any kind of teleology, because necessarily unintentional events produced at the level of genes generate a variety of phenotypes that reproduce themselves in increasing quantity if they turn out to be well adapted to their environment. This powerful mode of argumentation allows us to explain the genesis of plant and animal species without invoking a Pre-existing Designer and it has opened new doors for understanding a number of phenomena about living beings and societies including naturally human beings and human societies. Without a doubt, this includes the social framework developed over centuries in which economic phenomena take place. But when it comes to explaining phenomena that clearly result from a set of conflicting intentional and relatively rational decisions, the relevance of natural selection is much less evident.

7But if I make room for other kinds of explanation in economics, why then did I focus almost exclusively on explanations based on rationality? I might simply answer that the very fact that this kind of explanation was attacked so many times in the past few decades was in itself a reason to revisit and evaluate it. I might also say that, whereas natural selection was first developed in biology, behavioural explanations in psychology and formal models in physics before being applied to economics, explanations based on rationality were mostly developed in economics. The history of their development was intimately linked to the history of economic thought before they invaded other areas, as recalled by Dumouchel himself. These facts justify writing a book on the theoretical problems related to this kind of explanation. As for natural selection, it was in no way my “target,” but I found important to discuss its relation to rationality, before putting in relation in my last chapter rational explanations with some theories of explanation in the philosophy of science that can help to characterize them.

8The remaining question concerns the future of this kind of explanation. Since I am little inclined to make such predictions, I could not exclude that the dim perspectives raised by Dumouchel will materialize. But let us have a closer look at the reasons invoked. Regarding the question of instrumental rationality, Dumouchel is right to emphasize both the importance of other kinds of rationality and the fact that learning may generate new ends as well as new preferences. If I nonetheless referred only to instrumental rationality, it is because I considered only the explanatory power of the notion of rationality. Let me attempt to explain what I mean with an example. If a woman who used to dedicate herself to make as much money as possible becomes aware of the needs of poor people and ends up adopting a new end which induces her to donate important sums of money to charity instead of investing it profitably, one can say that this is perfectly rational from a value rationality point of view, but can we invoke this kind of rationality to explain it? If the point is to explain why this person donates so much money to a charitable organization, the explanation can be that this donation is the rational thing to do given her new end, but then the rationality invoked would be typically instrumental. If the point is to explain instead why she chose donating rather than investing, it might be sensible to observe that helping people is a higher value than making more money, but this perfectly value-rational answer does not convey any explanatory power. Explaining that A was preferred to B by claiming that A is more valuable than B does not explain very much, indeed. However, in spite of this lack of explanatory power, it is far from being excluded that, in the global explanation of complex economic phenomena, some room could be made for considerations about values.

9Regarding the role of the models as opposed to the role of rationality, one should not forget that the models themselves can be built up with rationality. This is at least the case with the famous neoclassical model I referred to in the passage on page 107 quoted by Dumouchel, which claims that profit is maximised at a point where marginal cost equals marginal revenue. This elegant geometric and algebraic model is nothing but an expression of the idea that entrepreneurs should be rational enough to keep up with production if extra gains are superior to extra costs and stop production when this trend is reversed. In such cases, the reconstruction of the situation with models can hardly be opposed to the “the ‘principle’ which ‘animates’ the explanation.” One should also note that this principle should be said to be false only when it is taken in its maximizing and universal version referred to by Popper, but not, as I recommend, in its minimal version. The fact that drivers sometimes do irrational things does not falsify the idea that people usually make sensible decisions. Maximisation is an idealisation of the real situation, but as Dumouchel says, it is “a methodological recommendation that guides us in the analysis of the situation in which the agents act,” a dimension that I have also, although briefly, evoked (p. 42). Therefore, such models hide but do not destroy the fundamental role of the rationality principle.

10These observations will not warrant a safe future for this principle, but I never claimed that such a guarantee could be provided. In my book, I rather analyzed how this principle works and how its continuous role has been hidden and forgotten. I then stated why, in spite of the significance of the theoretical developments such as those aptly referred to by Dumouchel, I remained unconvinced by the arguments in favor of its dismissal.

Haut de page

Pour citer cet article

Référence papier

Maurice Lagueux, « A Reply to Paul Dumouchel », Œconomia, 2-2 | 2012, 247-251.

Référence électronique

Maurice Lagueux, « A Reply to Paul Dumouchel », Œconomia [En ligne], 2-2 | 2012, mis en ligne le 04 juillet 2015, consulté le 23 juin 2017. URL :

Haut de page


Maurice Lagueux

Université de Montréal

Haut de page

Droits d’auteur

Licence Creative Commons
Les contenus d’Œconomia sont mis à disposition selon les termes de la Licence Creative Commons Attribution - Pas d'Utilisation Commerciale - Pas de Modification 4.0 International.

Haut de page
  • Logo Association Œconomia
  • Logo CNRS
  • Les cahiers de