Projectual Abduction - RERO DOC

Projectual Abduction GIOVANNI TUZET, University of Lausanne. Lausanne (Switzerland) and University of Ferrara. Ferrara (Italy). Email: [email protected] Abstract Projectual abduction is the inference drawing the means to achieve an end. Planning a course of action is an inferential task and we claim that the relevant inference is abduction. We distinguish projectual abduction from epistemic abduction. While epistemic abduction aims to determine an explanatory relation, projectual abduction aims to determine a teleological relation. It is important to remind in any case that abduction does not stand by itself: as is true for epistemic abduction, projectual abduction has to be developed and evaluated by subsequent deduction and induction. After defining projectual abduction in section 1, we focus on the relations between projectual abduction, normativity and truth in section 2. Then we sketch how projectual abduction works in artistic creativity (section 3), in evolutionary and teleological processes (section 4) and finally in social processes (section 5). Keywords: abduction, practical reasoning, normativity. Ma conviction, d` es la jeunesse, fut que, dans la phase la plus vivante de la recherche intellectuelle, il n’y a pas de diff´ erence, autre que nominale, entre les manœuvres int´ erieures d’un artiste ou po` ete, et celles d’un savant. (Paul Val´ ery)

1

Projectual Abduction

Projectual abduction is the inference drawing the means to achieve an end. Planning a course of action is an inferential task: we claim that the relevant inference is abduction. This is not the usual sense in which abduction is studied. Abduction is usually taken as an explanatory inference. Indeed Charles Sanders Peirce calls abduction the inference suggesting an explanation, notably (but not necessarily) a causal explanation. Sometimes he calls it retroduction. Such an inference is to be articulated with deduction and induction. According to him, their articulation has not only a logical character, but also a methodological value. We claim that such an articulation holds not only for explanatory tasks, but also for projectual tasks. That is, not only for the discovery of actual causes explaining an observed effect (or a law explaining an observed phenomenon) but also for the discovery of possible means to achieve an and. This is not the place to recall in detail the steps leading to Peirce’s theory of abduction. Let us say just the following before we focus on projectual abduction. In his early writings Peirce does not speak of abduction but rather of hypothesis, distinguishing hypothesis and induction, which are ampliative inferences, from deduction, which is non-ampliative – see e.g. [23, vol. 1, pp. 180, 266 ff., 362, 430 ff., vol. 3, pp. 323–338]. The point being the difference between ampliative and non-ampliative inferences, he does not always draw in his early writings a clear distinction between hypothesis and induction. Yet he says that induction determines a general character while hypothesis permits the knowledge of causes [23, vol. 1, p. 428](cf. e.g. [22, 5.272-276, 2.624]). In his later writings he makes a threefold distinction between abduction, deduction and induction – see e.g. [22, 7.162-255]; [22, 5.14212]. This threefold distinction is not only a logical one, but also a methodological one c The Author, 2006. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected] doi:10.1093/jigpal/jzk011

152


expressing the principles of scientific inquiry. According to Peirce every scientific inquiry is constituted by three inferential steps: first, by abduction we suggest a hypothesis explaining a fact; second, by deduction we determine the conceivable consequences of the hypothesis; third, by induction we test the conceivable consequences of the hypothesis verifying whether or not they correspond to real consequences.1 Peirce stresses that abduction is the only inference producing new concepts and theories [22, 2.96, 2.777, 5.171, 7.219]. However different conceptions of abduction have been elaborated by scholars working on it. Eco [5] classifies abductions according to their degrees of creativity. Thagard [29] classifies abductions depending on what is being abduced. Magnani [21] distinguishes in particular manipulative from theoretical abduction. Now what about projectual abduction? We said projectual abduction is the inference drawing the means to achieve an end. This is an informal representation of the reasoning in question:

How to get E? Tentatively, if C then E. Therefore, we should try with C.

Being C a cause (an intentional cause or a natural cause that can be provoked intentionally),2 C is abduced in order to attain E. The inference starts from an effect E to be attained and concludes to the cause C which is supposed to provide it. The conditional “if C then E” (where C is a means to attain E) we shall call practical conditional. Depending on such a conditional, the related conditional “if you want E you have to produce C” we shall call final conditional. In the next section we will show the relations between projectual abduction, normativity and truth. In sections 3-5 we will specify the modalities of projectual abduction and the inferential relations it has with deduction and induction. A preliminary remark to be made concerns the name of such an inference. Why do we choose projectual abduction? It could have been otherwise projective abduction. But the word “projective” bears some meaning which is for our purposes misleading: in particular it suggests the idea of transferring something actual to non-actual situations or the idea of deriving non-actual things (for instance possibilities) from actual things (Goodman [10, part II]). Otherwise it could have been planning abduction or practical abduction, but these names would have been too strict: “planning” seems to refer properly to the intellectual side of a course of action; on the other hand “practical” seems to focus more on the active than on the intellectual side of it; “projectual” seems to be a good compromise and a sufficiently general term. We take projectual abduction to be the general name of the inferences drawing means for intended ends. It gives an account of the first inferential step in tasks like planning, engineering, designing technical artifacts, creating a work of art. Designers, for instance, often describe their own activity as starting with functional specifications and ending with structural descriptions. It makes sense to think they explain their designs by describing post 1 Cf. especially Fann [7]. On the logical and semiotic models involved in Peirce’s theory of inference and scientific method see Tiercelin [30]. On abduction and induction cf. Flach and Kakas [8]. On the distinction between explanatory and classificatory abductions (notably in legal reasoning) cf. Tuzet [31]. 2 For the present purposes we assume that a means-end relation is a type of causal relation. The point is disputable but the inferential status of projectual abduction does not depend on that.

Projectual Abduction 153 hoc how the physical object can fulfill the function. These are descriptions of successful projectual abductions. Let us remind on the other hand that abduction is a necessary step but not a sufficient one. Planning, engineering, designing, creating are started and managed by a reasoning process which is self-correcting. To avoid errors and revise plans, projectual abductions must be followed by other inferences determining the conceivable consequences of projectual hypotheses and verifying their actual consequences. The scientific method we sketched above, constituted by the inferential articulation of abduction, deduction and induction, holds not only for explanatory tasks but also for projectual tasks.

2

Projectual Abduction and Normativity

Projectual abduction has a peculiar normative character. It differs from other kinds of normativity because it depends on the existence of an empirically verifiable relation, that is a means-end relation. Its normative status is not strictly speaking a deontic status. The norm expressed by a projectual relation is rather a “technical norm”, that is a norm prescribing some definite means to achieve a definite end (cf. [34, 2]). Let us call practical conditional the conditional (1) expressing a means-end relation. (1) If you take m, you will recover from d. Then, we shall call final conditional the conditional (2) expressing a normative relation depending on (1) and having the same truth-value of (1). (2) If you want to recover from d, you have to take m. Projectual abduction infers the assumption of medicine m from the intention of recovering from disease d. Such an inference and the related course of action will be successful if conditional (1) is true. From a different perspective, a subject A observing another subject B taking m, might infer the intention of B to recover from d. In this case, the condition for the inference being true is not the truth-value of (1) but the intentional state of B (which is not necessarily correspondent to (2) – it could be case that B takes m for other reasons). This second perspective is clearly epistemic, while the first is practical and projectual. The late von Wright [35, essay 5] distinguishes between norms and propositions of practical necessity related to norms.3 He claims that norms have a sort of primacy on propositions of practical necessity, since norms (which are neither true nor false) might remain unchanged while the related propositions (which are true or false) might change depending on the circumstances. (This is probably the reason why in 1996 he prefers to call them “propositions of practical necessity” instead of “technical norms”.4 ) He believes this account credits truth with a more important role in normative contexts. For not only the existential propositions (“norm-propositions”) to the effect that such and such norms exist (have been given) are true or false, but also the statements about what must or may be done in order to follow their rule [35, p. 70]. This is quite correct, but a further step might be made if we recognize that a norm prescribing a practically unattainable state of things is no norm at all, but just 3 “Norms primarily prescribe what ought to or may be; propositions of practical necessity state what the addressees of norms must or may do in order to satisfy their obligations or avail themselves of their permissions” [35, p. 68]. 4 “The statements that the agent must or must not do this or that in order to satisfy his obligation are true (or maybe false) in relation to the norm imposing the obligation. These statements of practical necessity should therefore better not be called themselves norms”[35, p. 69].

154


a senseless command. If this is true, practical attainability is a necessary condition of norms and projectual abduction is a necessary inferential means to determine norms. A challenge to projectual abduction might be the following. If it is well-known that taking m recovers from d the inference concluding that you have to take m if you want to recover from d is not a creative one. This seems to be at odds with the claim that abduction is the only inference producing new ideas. In fact it should be noted that in common-sense reasoning and in ordinary reasoning in general abduction has a less ambitious character. Now our intuition is that various kinds of abduction (cf. Eco [5], Thagard [29]) could be roughly summarized in the distinction between ordinary abduction and extraordinary abduction. The first obtains in common-sense explanations or in any explanation using already-known concepts and rules. Where, to say it differently, what is to be inferred is an unknown token of a known type. In the second instead what is to be inferred is an unknown token of an unknown type, when current cognitions do not permit a satisfying explanation of the case in hand – these are the cases Peirce calls “surprising” [22, 5.189, 7.220]. Obviously the second kind of abduction is more risky, but more fruitful when true, since extraordinary abduction introduces new hypotheses. This distinction holds for both epistemic abduction and projectual abduction: an ordinary projectual abduction deals with an already-known means-end relation, while extraordinary projectual abduction guesses at a new means-end relation. The question may be put whether being extraordinary is a necessary condition for being a creative abduction.5 The answer is in the affirmative if creation is relative to a new type of means-end relation (or explanatory relation in case of epistemic abduction); it is in the negative if creation is relative to tokens as well, for an ordinary abduction infers an unknown token of a known type. But of course the issue of creativity is much more interesting when relative to types. In the next section we will see how projectual abduction works in artistic creativity. But in a rather particular sense, as we shall see in section 4, it also bears on evolutionary and teleological processes. Before going into that, let us consider another challenge to projectual abduction: since in general there are lots of means to achieve an end, including implausible means, projectual abduction allows lots of conclusions, including implausible ones.6 Consider the following case:

I want to reach the airport (E). If I take a Limousine (C), I will reach the airport (E). Therefore, I have to take a Limousine (C).

In what sense a Limousine is needed? I could take a bus as well, or a Ferrari. This is true, but plainly our practical inferences are always performed under some constraints on means selection. Consider the next example taken form John Searle [28, p. 244]:

I want this subway to be less crowded (E). 5

We would like to thank an anonymous referee for this question.

6

Thanks to Professor Pierdaniele Giaretta for putting our attention on that.

Projectual Abduction 155 If I kill all the other passengers (C), it will be less crowded (E). Therefore, I have to kill all the other passengers (C).

Again, it is to be noted that our practical inferences are always performed under some practical constraints. So our reply to the present objection is this: economical or moral constraints always play a determinant role in projectual abduction. Note that similar constraints hold for epistemic abductions as well (see Peirce on the “economy of research”, [22, 7.140]). However the objection could be stated differently: claiming that projectual abduction provides solutions for practical issues begs the question because the main problem is the establishing of the premises, not the drawing of the conclusion. Then our reply would be simple: that is not a problem specific to abduction; for any inference concerns stricto sensu the drawing of conclusions from established premises, not the establishing of the premises. Projectual abduction is the inference drawing the means for intended ends under some practical constraints. Both in case the ends are inferred and in case they are not, their means are abductively inferred.

3

Projectual Abduction and Creativity

Douglas Anderson [1] claimed that Peirce has an explicit theory of scientific creativity and an implicit theory of artistic creativity. Besides his scientific methodology an analogous artistic methodology can be seen. As the scientific one, it is constituted by three inferential steps: artistic abduction, artistic deduction, artistic induction. In the first step, artistic abduction, a combination of ideas occurring mostly by chance indicates to the artist the way to satisfy an indeterminate telos he began with, such a telos being the expression of a certain feeling in which the sense of the work of art consists. But artistic abduction is perfectly fallible and not enough to achieve the artist’s goal. Thus he must trust in a subsequent artistic deduction elaborating and improving the initial project. In performing this task artistic deduction excludes from the project other less satisfying ways to achieve the goal and determines the telos with a greater precision. Then the artist entrust to artistic induction, to the testing of the results he obtained. (Think of a musical melody and its variations, played many times after being composed, so as to check them at different tempos and volumes). To sum up: “the telos arises in artistic abduction, artistic deduction projects for and precedes the telos, and artistic induction tests the finished work of art for esthetic value” [1, p. 151]. Artistic induction notably develops a self-correcting process, in order to find the more satisfying way to achieve the telos the work is created for; in other words, to find the best way of expressing the feeling the work wants to communicate. “Thus, the Peircean artist is like a scientist in employing three modes of mental activity in pursuit of an end” [1, p. 150].7 With a fundamental difference: scientific abduction aims to determine the actual causes of observed phenomena (or the laws explaining them), while artistic abduction aims to determine the best means for intended ends (the expression of a feeling, of a sense). In other words (cf. [11, 14]), scientific abduction aims at the best explanation of a fact, and artistic abduction at the best realization of an esthetic end. In sum we should express the difference in these terms: scientific abduction has an epis7 No doubt in these processes the imaginative capacities of intelligence are at work. “People who build castles in the air do not, for the most part, accomplish much, it is true; but every man who does accomplish great things is given to building elaborate castles in the air and then painfully copying them on solid ground. Indeed, the whole business of ratiocination, and all that makes us intellectual beings, is performed in imagination” [22, 6.286].

156


temic character, while artistic abduction has a projectual character. The result of an epistemic abduction is a knowing-that. The result of a projectual abduction is a knowing-how. The distinction between ordinary and extraordinary abduction sketched in section 2 applies here as well. An ordinary projectual abduction deals with an already-known means-end relation, while an extraordinary projectual abduction guesses at a new means-end relation. Consider the following examples: for a Renaissance composer, variations on Madrigals were variations on an already-known artistic means; for Arnold Schoenberg, the 12-notes serial method was an absolutely new means for composing complex instrumental structures. The second is an example of a higher form of creativity. Anderson’s work is certainly important and remarkable. Yet we should try to integrate it, to turn out the general bearing of his insights. In fact, making abstraction from the peculiarities of any specific field, the methodology he develops applies to any creative and projectual activity (cf. [4]). To any form of activity in which an intelligence seeks the means achieving an end, in other terms the possible causes producing a practical effect (or in even other terms the possible efficient causes satisfying a final cause). Once the means is abduced, its possible consequences must be drawn in order to verify in principle whether or not they can achieve the intended end (at the same time the end itself is specified). Such a drawing of the consequences is the deductive phase of the reasoning process. Then, by realizing the project it can be verified whether the work satisfies the end: it is the inductive phase, the testing of the work. This projectual methodology applies to technological innovation, to artistic creation and intellectual creation in general, to commercial activities, etc. To any form of final action. (Political and legal action included, notably legislation. For instance: if a policy so-and-so will be implemented, the social values so-and-so will be realized.)

4

Projectual Abduction and Teleological Processes

In general we call projectual this kind of abduction: the one not delivering an explanation (as epistemic abduction does) but suggesting some means to achieve an end. We shall try to explain now why projectual abduction bears on evolutionary and teleological processes as well. In some places Peirce says that nature makes retroductions (cf. [24, p. 161]). What does he mean? He does not obviously mean that nature explains backwards her own processes, guessing at the causes of actual effects. He does mean that nature is capable of finding the means to achieve an end she pursues. This imply believing that nature operates in teleological terms, that evolutionary processes are not blind processes governed by chance, but goal-oriented processes. In a certain sense, if Peirce is right, nature operates a sort of planning, putting forward some goals and then selecting the best means to achieve them. Some scientists might object to this picture, claiming it is naive or unnecessary given that science can explain evolutionary processes without having recourse to teleological concepts. However this is not the fundamental point at issue here. What we do claim is that projectual abduction has a fundamental role in developing intelligence and determining adaptation. To put it very intuitively, creatures not capable of good projectual abductions have little chance of survival. Moreover, projectual abduction is fundamental in practical and social matters. Thus, if evolutionary processes are not the result of chance or blind necessity, one at least of the ways they are realized is projectual abduction by intelligent creatures or by nature itself if we accept a teleological picture of nature. The distinction between ordinary and extraordinary abduction applies also here. Taking medicine m to recover from disease d (when m is already-known) is an example of ordinary

Projectual Abduction 157 projectual abduction. Let us make instead a more complex example of extraordinary projectual abduction. Consider how the ram was invented according to Vitruvius’ De architectura (X, XIII). The Carthaginians besieging Cadiz wanted to demolish the defensive works. Since they hadn’t appropriate iron engines, they took a beam and started to beat repeatedly the walls: so they were capable of throwing down the higher stones first and demolish the entire defensive works in the end. This is an example of successful extraordinary projectual abduction. Epistemic and projectual abductions are of course related. To find abductively a good means for an end is to guess at the causal chain producing it. This could be more successfully performed by creatures also capable of guessing at the causes of the effects they observe, that is, capable of epistemic abductions. In general, projectual abduction cannot disregard the knowledge of the relevant facts; in this sense it has an epistemic dimension. To be justified, projectual conjectures must be based on adequate information and oriented towards sufficiently determined goals.8 As Hempel [13, p. 464] roughly puts it, an action is rational if, on the basis of the given information, it offers optimal perspectives of goal-achieving.9 And vice versa, to find abductively a good explanation is very often a practical task requiring the subject to move and operate successfully in the environment, that is making projectual abductions (cf. Maddalena [20] and Fabbrichesi Leo [6] on “practical retroduction”). Moreover, the practice of explanation often requires manipulations and modifications of the initial conditions in order to select and test hypotheses (cf. Magnani [21] on the notion of manipulative abduction). Has one of the two abductions – epistemic and projectual – a sort of cognitive primacy on the other? We think it is impossible to claim the one precede the other. For it would imply the awkward claim that practical knowledge is prior to theoretical knowledge or vice versa, which are both implausible claims, given the inextricable entanglement of knowing-how and knowing-that from which intelligence develops. Now, does an inferential and teleological account like the present provide an explanation (or at least a partial explanation) of practical reason? When Searle ([28, chap. 8]) denies there is a deductive logic of practical reason, he is right; but not because there is no logic at all of practical reason: in fact there is an abductive logic of practical reason (better, the first step of practical reasoning is abduction). Searle seems to say that means-end reasoning has neither validity nor plausibility as an explanation of practical reason [28, pp. 242–248]. On validity we would agree if reasoning were just deductive reasoning: in fact reasoning is not just deductive reasoning and validity is not just deductive validity; abductive and inductive reasoning have their own criteria of validity [18]. On plausibility the problem is the following: given an end, there is an indefinite number of means, many of them implausible, to achieve it – that is, an indefinite number of means constituting sufficient conditions for it (cf. the previous objection of the Limousine); so the reasoning would be sound only in case of necessary conditions. But, as we said before, our practical inferences are always performed under some constraints on means selection. Those constraints can be classified in two main groups, (i) economical and (ii) moral constraints. The problem is not to find a means whatsoever, but the best means relatively to the constraints and the context (Searle is 8 On the rationality of action according to the “information-basis” and the “objectives of an action” cf. e.g. Hempel [13, pp. 463–469]. Cf. Harman [12, chaps. 8-9], Bratman [3], Searle [28]. 9 More precisely (Hempel [13, p. 465]): “if the information basis contains general laws by virtue of which certain of the available actions would be bound to achieve the total objective, then, clearly, any one of those actions will count as rational in the given context. If the information basis does not single out any available action as a sufficient means for attaining the objective, it may yet assign a numerical probability of success to each of the different available actions; in this case, any action will count as rational whose probability of success is not exceeded by that of any available alternative.” As Hempel stresses this is a simplified model; on choice in conditions of uncertainty cf. [13, pp. 466–469].

158


clearly aware of that, but he doesn’t draw the due consequences of the point). The relevant inference is – to draw a parallel with Inference to the Best Explanation – Inference to the Best Solution.10 Thus, such an objection as Searle’s does not constitute an objection to the claim that an inferential and teleological account as ours constitutes an explanation (at least a partial explanation) of practical reason.

5

Projectual Abduction and Social Processes

Projectual abduction has an important role in social matters as well. It is the inferential means to find the ways of achieving social goals. Economical projectual abductions and legal projectual abductions are remarkable examples of that. We shall say something on legal projectual abductions. We won’t discuss the peculiar modalities of finding, elaborating and evaluating a socially relevant projectual hypothesis. There is work to be made on that, on the peculiar conditions of testing and accepting a socially relevant projectual abduction; but perhaps such a work does not bear on a set of conceptual problems involving methodological differences with other kinds of projectual abductions: rather it bears on the sociological aspects of the question; in the sense that what is at stake are not the logical and methodological conditions of socially relevant projectual abductions but rather the conditions of their acceptability in a social context. This being said, let us review some kinds of legal projectual abductions. The lawyer’s, the judge’s and the legislator’s projectual abductions shall be sketched. (For each of them there are of course ordinary and extraordinary forms, but notice that in the legal domain prudence is usually more important than creativity). The lawyer ’s method involves a kind of projectual abduction: the lawyer determines a result he wants to achieve and then seeks the best means to attain it. He seeks the best means to support his thesis, and the best strategy to persuade the jury or the judge that his thesis provides the best explanation of the facts to be judged and the best solution for the case in hand. A similar discourse can be made concerning the judge. The outcome of a legal process can be determined by a judicial inferential process starting with a projectual abduction. The judge selects a social value to be realized, seeks for the best means to attain it and then decides the case consequently. But an important distinction shall be made here. In continental legal systems the outcome should follow in general from normative premises already established by legislation; in these terms there is little room for projectual inference, since the judge is expected to draw conclusions from premises already established (cf. Tuzet [32]). In common law systems, on the contrary, judges have a broader power of selecting the outcomes. Case-based reasoning and teleological reasoning are typical forms of common law judicial reasoning (cf. Levi [17]; Sartor [27]). Continental systems are intended to satisfy objectivity and impartiality but are lacking in flexibility. Common law systems put a great value in flexibility and revision but are lacking in certainty and uniformity. This is not the place to compare the advantages and drawbacks of the systems; what matters is whether projectual reasoning has a role in judicial reasoning: at least in common law systems it has a role in the selection of the outcome (means) given a social value (end). Finally, in any legal system the legislator has recourse to projectual reasoning and abduction in order to determine goal-oriented legislation. When statute law is intended to satisfy 10 There is a clear analogy with what van den Bosch ([33, pp. 489-490]), in the domain of philosophy of medicine, calls Inference to the Best Manipulation. We think that Inference to the Best Manipulation is a specific case of Inference to the Best Solution.

Projectual Abduction 159 some social goals and values, it must consist in legal norms providing the best means to achieve the goals. (Obviously not the best means in absolute but relatively to the context, to the information and the constraints eventually imposed on the task). Projectual abduction, it should be reminded, differs from epistemic abduction because the latter starts from an actual result, while the former starts from an hoped result the means of which are sought. While epistemic abduction aims to determine an explanatory relation, projectual abduction aims to determine a teleological relation.11 In these terms, no doubt the scientific method we sketched applies to the so-called science of legislation: first, by projectual abduction the legislator tries to find the best means for a social and legal end; secondly, by projectual deduction the foreseeable consequences of the application of the means are drawn; thirdly, by an inductive process the actual consequences of the means are collected and evaluated, verifying whether or not they satisfy the intended end. Of course the application of a scientific methodology to social matters does not guarantee the achieving of the goals. For the simple reason that the goals might change before the method brings to successful results. In fact, truth being the invariable goal of science, Peirce claims that the scientific method will bring to truth in the long run: a sufficiently long process of revision will discriminate true from false beliefs.12 But social values might change before a process has achieved them. On the one hand, their deciduous nature prevents from their achievement. On the other, their fixation is not a price to pay: freedom in selecting values has no less importance than method in realizing them. In any case, we believe, projectual reasoning is an indispensable means for achieving or just approximating social values.

References [1] D. A. Anderson. Creativity and the Philosophy of C. S. Peirce. Nijhoff, Dordrecht, 1987. [2] G. M. Azzoni. Cognitivo e normativo: il paradosso delle regole tecniche. Franco Angeli, Milan, 1991. [3] M. E. Bratman. Intention, Plans, and Practical Reason. Harvard University Press, Cambridge, MA, 1987. [4] V. M. Colapietro. Portrait of an historicist: an alternative reading of peircean semiotic. Semiotiche, 2:49–68, 2004. [5] U. Eco. Horns, hooves, insteps: some hypotheses on three types of abduction. In U. Eco and T. A. Sebeok, editors, The Sign of Three: Holmes, Dupin, Peirce, pages 198–220. Indiana University Press, Bloomington, 1983. [6] R. Fabbrichesi Leo. L’abduzione come “profezia retrospettiva”. Semiotiche, 2:123–135, 2004. [7] K. T. Fann. Peirce’s Theory of Abduction. Martinus Nijhoff, The Hague, 1970. [8] P. Flach and A. Kakas, editors. Abductive and Inductive Reasoning: Essays on Their Relation and Integration, Dordrecht, 2000. Kluwer Academic Publishers. [9] J. Frank. Courts on Trial. Myth and Reality in American Justice. Princeton University Press, Princeton, 1949. [10] N. Goodman. Fact, Fiction, and Forecast. Harvard University Press, Cambridge, MA, 1954. [11] G. Harman. Inference to the best explanation. The Philosophical Review, 74:88–95, 1965. [12] G. Harman. Change in View. Principles of Reasoning. The MIT Press, Cambridge, MA, 1986. [13] C. G. Hempel. Aspects of Scientific Explanation. And Other Essays in the Philosophy of Science. The Free Press, New York, 1965. 11 On law as a rational teleological enterprise see e.g. MacCormick ([19, p. 125]). “Any law which is not expected to operate in the future is not evidence of a functioning law, but is a definition only” ([15, p. 27]). On the topic of the social consequences of the law the literature is quite abundant; see for instance an author belonging to the so-called Sociological Jurisprudence as Pound [26]; cf. Frank [9, chap. 4], Posner [25]. 12

This is a basically realist claim, even though it is susceptible of an instrumentalist interpretation. See Kuipers [16].

160


[14] J. Josephson and S. G. Josephson. Abductive Inference. Computation, Philosophy, Technology. Cambridge University Press, Cambridge, 1994. [15] R. Kevelson. Peirce, Paradox, Praxis. Mouton de Gruyter, Berlin and New York, 1990. [16] T. A. F. Kuipers. Abduction aiming at empirical progress or even truth approximation leading to a challenge for computational modelling. Foundations of Science, 4:307–323, 1999. [17] E. H. Levi. An introduction to legal reasoning. The University of Chicago Law Review, 15:501–574, 1948. [18] I. Levi. Inference and logic according to Peirce. In J. Brunning and P. Forster, editors, The Rule of Reason. The Philosophy of Charles Sanders Peirce, pages 34–56. University of Toronto Press, Toronto, 1997. [19] N. MacCormick. Legal Reasoning and Legal Theory. Clarendon Press, Oxford, 1978. [20] G. Maddalena. Istinto razionale. Studi sulla semiotica dell’ultimo Peirce. Trauben, Torino, 2003. [21] L. Magnani. Abduction, Reason, and Science. Processes of Discovery and Explanation. Kluwer Academic/Plenum Publishers, New York, 2001. [22] C. S. Peirce. Collected Papers of Charles Sanders Peirce. Harvard University Press, Cambridge, MA, 1931-1958. vols. 1-6, Hartshorne, C. and Weiss, P., eds.; vols. 7-8, Burks, A., ed. [23] C. S. Peirce. Writings of C.S. Peirce: a Chronological Edition. Indiana University Press, Bloomington, IN, 1982. 6 vols. published, ed. by M. Fisch et al. [24] C. S. Peirce. Reasoning and the Logic of Things. Harvard University Press, Cambridge, MA, 1992. Edited by K. L. Ketner. [25] R. A. Posner. Law, Pragmatism, and Democracy. Harvard University Press, Cambridge, MA, 2003. [26] R. Pound. Mechanical jurisprudence. Columbia Law Review, 8:605–623, 1908. [27] G. Sartor. Teleological arguments and theory–based dialectics. Artificial Intelligence and Law, 10:95– 112, 2002. [28] J. R. Searle. Rationality in Action. The MIT Press, Cambridge, MA, 2001. [29] P. Thagard. Computational Philosophy of Science. The MIT Press, Cambridge, MA, 1988. ´ ´ [30] C. Tiercelin. La pens´ ee–signe. Etudes sur C. S. Peirce. Editions Jacqueline Chambon, Nˆımes, 1993. [31] G. Tuzet. Legal abductions. In D. Bourcier, editor, Legal Knowledge and Information Systems: Jurix 2003, pages 41–49, Amsterdam, 2003. IOS Press. [32] G. Tuzet. Abduction and deduction: epistemic justice vs. political justice? In C. Dahlman and W. Krawietz, editors, Values, Rights and Duties in Legal and Philosophical Discourse, pages 211–221, Berlin, 2005. Duncker & Humblot. [33] A. P. M. van den Bosch. Inference to the best manipulation - a case study of qualitative reasoning in neuropharmacy. Foundations of Science, 4:483–495, 1999. [34] G.H. von Wright. Norm and Action. Routledge and Kegan Paul, London, 1963. [35] G.H. von Wright. Six essays in philosophical logic. Acta Philosophica Fennica, 60:5–91, 1996.

Received July 31, 2005