(Forthcoming in Kant-Studien) Kants Non-Teleological Conception of Purposiveness by Thomas Teufel, New York Abstract In this paper I argue, first,

that Kants technical definition of purposiveness in 10 of the Critique of the Power of Judgment is designed to abstract from all forward-looking considerations (teleological, intentional, normative) that accompany the conventional understanding of the term. Kant seeks to establish a strictly backward-looking, etiological notion of purposiveness in order to capture the causal link connecting artifacts with their concepts. I argue, second, that he succeeds. Kants etiological conception of purposiveness neither collapses into mechanism, nor smuggles normative considerations in through the backdoor. I frame my discussion by critically engaging Hannah Ginsborgs reading of 10the leading example of a normative interpretation of Kants notion of purposiveness. Keywords: purposiveness, conceptual causality, normativity, Hannah Ginsborg. 1. Introduction A perennial task for commentators on Kants Critique of the Power of Judgment is to explain what Kant means by purposiveness. Kants notion of purposiveness is the central term of art in the third Critique: it pertains to virtually all of the books major doctrines and, in consequence, holds the key, also, to the vexing problem of the books unity. In spite of this centrality, the notion of purposiveness remains poorly understood. This is not altogether surprising. The meaning philosophers associate most closely with the notion of purposivenessits teleological meaningis just what Kant hopes to do without. Not only is teleology not an important component part of Kants conception of purposiveness, Kant seeks to bar it from his technical definition of the term altogether.1 This has not been appreciated in the literature: the highly instructive route by which Kant arrives at his definition of purposiveness has remained obscure as has, accordingly, the nature of the technical vocabulary it yields.2
1

This non-teleological conception of purposiveness is not to be confused with Kants doctrine of a purposiveness without purpose in the Critique of the Aesthetic Power of Judgment. The various modifications of purposiveness that Kant proposes over the course of the third Critiqueincluding the one in the Aestheticpresuppose Kants base non-teleological conception. Commentators, for the most part, do note the peculiar etiological nature of Kants technical conception of purposiveness that I highlight below (see 1.1 and 2.3) but then either fail to note its corresponding nonteleological nature (see 2.3 and 2.5) or, at any rate, to prevent conventional teleological considerations from encroaching on Kants technical conception: Konrad Marc-Wogau (Vier Studien zu Kants Kritik der Urteilskraft. Uppsala. 1938) acknowledges Kants etiological meaning of purposiveness butdriven

Following introductory remarks (see 1.1-1.2), I advocate two related claims. First, I argue, on textual grounds, that Kant seeks to establish a non-teleological, etiological conception of purposiveness as his central technical term in the third Critique (see 2.3). Second, I argue, on philosophical grounds, that this non-teleological conception of purposiveness is successful, and neither collapses into mere mechanism (see 2.4), nor smuggles teleology in through the backdoor (see 2.5). I frame my discussion by critically assessing Hannah Ginsborgs normative interpretation of Kants notion of purposiveness.3 Normativity is the most resilient element of teleological interpretations of Kants notion of purposiveness, and Ginsborgs interpretation is, for that reason, an important test case for my own (see 2.1, 2.2, and 2.5).
by his systematic ambition to uncover the dialectical nature of Kants conception of purposiveness presents this etiological meaning as inextricably linked to means-end considerations (47ff.); Giorgio Tonelli (Von den Verschiedenen Bedeutungen des Wortes Zweckmigkeit. Kant-Studien 49. 195758) complains at length about the shifting meanings of purposiveness in the third Critique but takes a teleological core notion, however variously modified, for granted throughout; Klaus Dsing (Die Teleologie in Kants Weltbegriff. Bonn. 1968) characterizes Kants notion of conceptual causality as, fundamentally, one of [] teleological causality [] (97); John McFarland (Kants Concept of Teleology. Edinburgh. 1970) initially distinguishes Kants etiological from the conventional teleological meaning (78) but subsequently ignores that distinction (80ff.); Helga Mertens (Kommentar zur Ersten Einleitung in die Kritik der Urteilskraft. Mnchen. 1975) explains the etiological meaning correctly (109) but nevertheless continues to characterize Kants formal-logical as well as his real-absolute purposiveness in teleological terms (107, 111); Paul Guyer (Kant and the Claims of Taste. Cambridge. 1979) distinguishes between Kants etiological and the conventional teleological notion of purposiveness with great clarity but rejects the relevance of the former for Kants Aesthetic. (188 ff.); Clark Zumbach (The Transcendent Science. The Hague. 1984) does not distinguish between Kants etiological and the conventional teleological meaning of purposiveness but instead between two teleological meanings: the conventional one and what he calls [] designedness [] (p. 7; roughly, between what is actually designed and what only appears designed); Christel Fricke (Kants Theorie des reinen Geschmacksurteils. Berlin. 1990) initially characterizes Kants etiological notion of purposiveness correctly in her account of 10 of the third Critique (72ff.) but then muddies the waters by identifying Kants notion with a [] hypothetical purposiveness [] (82) of her own invention that retains a teleological flavor as it continues to point to [] hypothetical ends [] (97); Hannah Ginsborg (Kant on Aesthetic and Biological Purposiveness. In Andrews Reath, Barbara Herman, Christine Korsgaard (eds.): Reclaiming the History of Ethics: Essays for John Rawls. Cambridge. 1997) acknowledges Kants etiological conception of purposiveness but argues at length that this conception is normative, hence teleological, after all (329-60; see 1.2, 2.2, 2.5, below); Henry Allison (Kants Theory of Taste. Cambridge. 2001) acknowledges the etiological character of Kants technical conception of purposiveness (121) but then continues to claim that something is purposive in Kants technical sense if it [] enhances [] (127) something else, thus drawing on the conventional teleological conception; Peter McLaughlin (What Functions Explain. Cambridge. 2001) notes that Kant, in the third Critique, gives precedence to the etiological meaning of purposiveness (23n.; see also his Kants Kritik der Teleologischen Urteilskraft. Bonn. 1989, 42) but then goes on to use what he alleges to be Kants conception of purposiveness in an analysis of organisms as teleological systems having [] a good [] (209); Rachel Zuckert (Kant on Beauty and Biology. Cambridge. 2007) insists, with reference to Kants technical definition of purposiveness in 10, that Kants [] most basic understanding of [] purposivenessis intentional [] (79) and, hence, teleological. Ginsborg, op. cit.

1.1 Kants Definition of Purposiveness: An Overview Kant defines what he means by purposiveness with precision early on in the third Critique, in 10, entitled On purposiveness in general4. Kant there presents purposiveness as his terminus technicus for mind/world causality, and, specifically, for conceptual causality. Purposiveness, Kant says, is [] the causality of a concept with regard to its object []5. To characterize an object as exhibiting purposiveness (or as being purposive, or as a purpose), according to this definition, is to ascribe to the object a certain kind of causal ancestry. It is to make the etiological claim that the objects concept played a causal role in the objects coming into being, and, hence, in helping to shape and arrange the object as it is before us. In a first approximation, we may then say that to judge an object purposive, in Kants technical sense, is to detect the sort of order in it that can be imparted by a concept. Kants technical definition of purposiveness thus differs from the conventional understanding of the term. Ordinarily, to call an object purposive is to judge it to be directed at, useful for, well-suited to, or intended for an end. It is the reference to an end, or telos, in these characterizations that marks the conventional conception of purposiveness as teleological in the broadest sense.6 The idea of an origin in concepts (be it actual or merely apparent) may well resonate in the background of judgments of purposiveness in this conventional sense, but it hardly exhausts what we mean by them. If it is not immediately clear, then, just where the difference between Kants etiological notion of purposiveness (being caused by a concept) and the conventional teleological notion of purposiveness (being related to an end) may lie, it is telling that the unusual definition of 10 is Kants considered view of purposiveness in the Critique of the Power of Judgment. The prominently placed definition is the only definition of

KU, AA 05: 219.30. Von der Zweckmigkeit berhaupt. All translations from Kants works follow (with occasional slight modifications) Guyer, Paul and Wood, Allen (eds.): The Cambridge Edition of the Works of Immanuel Kant. 17 vols. Cambridge. 1995 . KU, AA 05: 220.3-4. [] und die Causalitt eines Begriffs in Ansehung seines Objects ist die Zweckmigkeit ( forma finalis). In a more precise sense, only the relation of being directed at an end is teleological simpliciter. The other relations mentioned are further determinations (respectively: functional, valuational, and intentional) of that base relation. Unless otherwise indicated, I use the notion of teleology in the broad sense noted.

purposiveness Kant gives in the entire third Critique7,8, and Kant continuously refers to it throughout the rest of the text.9

1.2 Hannah Ginsborgs Normative Interpretation of Purposiveness In her Kant on Aesthetic and Biological Purposiveness,10 Hannah Ginsborg seeks to minimize the apparent difference between Kants technical conception of purposiveness and the conventional conception of purposiveness. On Ginsborgs view, Kants definition of purposiveness in 10 of the third Critique is inherently teleological, after allin spite of its etiological focus on concepts as causes. Ginsborg presents her two-pronged argument in terms of an alleged normativity implicit in Kants definition. Ginsborg contends, first, that the idea of an object caused by a concept entails that the object caused be intended for an end. It thus entails the normative idea that the object caused conform to the end pursued. She holds, second, that the idea of an object caused by a concept entails that the object caused itself be an end. This is registered in the normative demand that an object caused by a concept conform to the concept that caused it. The idea of an object caused by its concept is thus inseparable from the normative idea that the object caused be as itby the standard of (i) its intended end11 and (ii) its conceptual cause12ought to be. I do not think that Ginsborgs argument is successful. Her attempt to link Kants notion of conceptual causality to the normativity of intentional causality (or design) is not convincing; and while there undeniably is a way that any given product of conceptual causality is to be (making it an end of sorts), judging something to be a product of conceptual causality does not require reference to what way that isand, hence, does not
7

Kants occasional characterization of purposiveness in terms of lawfulness does not constitute an alternative definition of purposiveness but, rather, highlights the circumstance that the order detected in a judgment of purposiveness is conceptual order (see especially EEKU, AA 20: 217.27-28). 8 Kant does, however, on at least three occasions, define the conceptual cause of an objectas opposed to the object itselfas a purpose, thereby contradicting the terminology adopted in 10 (see KU, AA 05: 180.31-2; EEKU, AA 20: 232.16-7; GTP, AA 08: 181.13). But note that this terminological uncertainty about exactly what should be called purpose in the causality of a concept with regard to its object evidently presupposes (it does not present an alternative to) Kants definition of purposiveness as the causality of a concept with regard to its object. 9 See KU, AA 05: 177.20n., 180.131-4, 307.29-30, 367.1-3, 369.33-5, 372.31-3, 393.31, 408.4-6, 454.236; EEKU, AA 20: 196.18-20, 217.24-7, 230.22-4; see also KpV, AA 05: 9.2n. 10 Ginsborg, op. cit. 11 See 2.2, below. 12 See 2.5.1, below.

entail a normative assessment of the object in question. Contrary to Ginsborgs claims, Kants etiological conception of purposiveness emerges as a fully non-normative and non-teleological conception.

2. Normativity Ginsborg first detects the normative dimension of Kants conception of purposiveness in his account of the purposiveness of nature in the First Introduction to the third Critique. She contends that this same normative dimension also informs Kants technical definition of purposiveness in 10. It is not necessary, at this point, to ask whether Ginsborgs discovery of normativity in Kants First Introduction is an accurate reading of that text. All we need to determine in order to get going is what sort of normativity Ginsborg believes to be at work there. The remaining task for the paper will then be to see whether Ginsborgs conclusion that such normativity is [] already implicit in Kants initial notion of purposiveness [in 10]13, is correct.

2.1 Normativity in the First Introduction Ginsborg originally detects normativity in Kants characterization, in the First Introduction to the third Critique, of the principle of purposiveness as a principle concerned with the [] conformity to law of the contingent []14. Kant presents a simple example of this conformity to law in our explanation of the structure of the human eye. When explaining the structure of the eye, we do not rest content to say, merely, that this structure happens to coincide with optical laws that are consistent with the possibility of vision. Instead, we judge that these optical laws govern the eyes structure, that the structure is the way it is in order to enable vision, and, hence, that the eye [] was

13 14

Ginsborg, op. cit., 342. EEKU, AA 20: 217.27-8. Denn Zweckmigkeit ist eine Gesetzmigkeit des Zuflligen als eines solchen. The above translation of Gesetzmigkeit as conformity to law is Ginsborgs. I use it in this section in order to present Ginsborgs normative account of the First Introduction without unnecessary complications. Note, however, that this translation makes a normative interpretation of Kants Gesetzmigkeit des Zuflligenand of the associated conception of purposivenessvirtually inevitable. Since the normativity of Kants conception of purposiveness is, at the very least, debatable, the more neutral translation of Gesetzmigkeit as lawfulness, which leaves the matter open, is to be preferred. See also KU, AA 05: 184.2-6.

supposed to be suitable for seeing []15. Kant notes, however, that the principle of such non-accidental conformity between the eyes structure and natural laws [] could not have been drawn from experience [] since experience [] teaches us only what things are []16, and not how they should be. To detect non-accidental conformity to natural laws is to go beyond a merely descriptive conception of the laws in question, and to see them in a normative light. Thus, according to Ginsborg, what Kant means by the conformity to law of the contingent is really its [] conformity to normative law.17 Whatever the grounds of this attribution of normative force to natural laws may be, the laws, thus construed, now allow us to judge not merely that the contingent object is a certain way, but why (to what end) it is that way, and whether it is as itby the standard of (i) that end and (ii) those lawsought to be. Ginsborg concludes that, since Kants notion of a conformity to law of the contingent must be understood in these normative terms, and since Kant calls such conformity to law purposiveness18, Kants notion of purposiveness in the First Introduction to the third Critique is a normative notion.

2.2 Purposiveness as Conformity to an Intended End Ginsborgs presents two arguments intended to connect this normative interpretation of the notion of purposiveness to Kants official definition of purposiveness in 10 of the Critique of the Power of Judgment. The first of these proceeds in two phases.19 In the first phase, Ginsborg argues for a close link between Kants technical definition of purposiveness, on the one hand, and intentional causality or design, on the other. Since an object produced intentionally is an object produced with an end in mind, purposiveness as defined in 10 emerges as a teleological notion. In the second phase, Ginsborg contends that Kants technical notion of purposiveness, thus understood, must be given a normative interpretation.

15

EEKU, AA 20: 240.25-6. Nur vom Auge urtheile ich, da es zum Sehen hat tauglich seyn sollen []. 16 EEKU, AA 20: 240.16-7. [] welches [Princip] aus der Erfahrung (die da nur lehrt, was die Dinge sind) nicht hat gezogen werden knnen. 17 Ginsborg, op. cit., 339. 18 See EEKU, AA 20: 217.28, 228.16-7; KU, AA 05: 184.2-6, 404.27-8 19 Ginsborg, op. cit., 331ff. Ginsborgs second argument is the topic of 2.5, below.

Ginsborg begins by noting that when Kant defines purposiveness as the causality of a concept with regard to its object, he means to suggest that [] the concept of a thing or state of affairs can play a causal role in bringing it about []20. In other words, Kant means to suggest that a concept can make a causal [] contribution [] to the processes by which change takes place in the spatio-temporal world.21 Ginsborg does not discuss what grounds Kant may have for endorsing the possibility of such mind/world causality or how we are to imagine it or, indeed, whether the notion makes (Kantian) sense at all.22 Instead, Ginsborgs main concern is to connect Kants notion of conceptual causality (or purposiveness) to the idea of intentional causality (or design)presumably because this way normativity lies. She explains that a concept is causally efficacious:

[] whenever a thing or state of affairs is brought about by an intelligent agent in accordance with an intention or design. If a thing is intentionally produced, there is a concept of the thing in the mind of the agent which is antecedent to the things existence and which governs or determines the agents activity in producing it.23 Ginsborg here makes an empirical-psychological observation about the circumstances under which conceptual causality [] happens []24i.e., when and where we locate causally efficacious concepts, namely: antecedent to the existence of their objects and in an intending agents mind. Note, however, that this presents intentional causality as, at best, a sufficient (not a necessary) condition for conceptual causality. That is to say, from Ginsborgs observation that whenever there is intentional causality, there is conceptual causality, the converse (that whenever there is conceptual causality, there is intentional causality) does not follow. Evidently, then, it would be a mistake to conclude that cases in which a concept plays a causal role in the production of its object (i.e., cases of purposiveness, in Kants sense) are automatically cases of intentional causality. Non-

20 21

Ibid., 331. Ibid. 22 See 2.4, below. 23 Ibid. 24 Ibid.

intentional conceptual causality remains a distinct logical possibility.25 Ginsborg, however, draws just this unwarranted conclusion, and takes her empirical-psychological observation to point not only to designs sufficiency for Kantian conceptual causality but also to something approaching necessity: [p]urposiveness, thus far defined, appears to be restricted to products of intentional causality.26 The very nature of the restriction to which Ginsborg points in her conclusion entails, of course, that even if Kants notion of conceptual causality were restricted to intentional causality (it is not),27 the notion of conceptual causality as such remains broader in scope. Ginsborg is aware of this, and, in the second phase of her initial take on 10, seeks to reinforce the conclusion that non-intentional conceptual causality cannot be what Kant has in mind in 10. It is here that the idea of normativity comes into play. Given the earlier shortcomings, one would expect an argument for the normativity of Kants notion of conceptual causality that is independent of the flawed attempt to connect puposiveness with design.28 Yet, the argument for the normativity of Kants notion of conceptual causality that Ginsborg now presents continues to depend upon the latter, unwarranted result. Ginsborg concedes that we can readily conceive of non-intentional conceptual causality. She imagines a painter in love who, rather than depict his actual model, always ends up painting the woman of his obsession29malgr lui30and explains that [h]ere the concept of the woman is causally efficacious in producing her likeness []31, even as it fails to influence the creation of her portrait intentionally. Ginsborg notes that such an unintended, sub-conscious conceptual influence is not, however, what we call design because it fails to be appropriately normative. As she puts it, [f]or the characteristic causality of design it is needed, not just that the concept influence the

25

Kant characterizes his conception of purposiveness in 10 as formal precisely in order to distinguish it from real or empirical purposivenessi.e., purposiveness as encountered in the actual production of artifactswhich, as a contingent matter of fact, presupposes intention. See, e.g., KU, AA 05: 364.23-5. 26 Ginsborg, op. cit., ibid., 332, my emphasis. 27 See 2.3, below. 28 As we will see, Ginsborg does have such an argument; see 2.5, below. 29 Ginsborg, op. cit., ibid., 342. 30 A less romantic version of this idea can be found in Emile Zolas Therse Raquin, where Laurent, aspiring artist and haunted murderer of his mistress husband, discovers, to his horror, that the features of his victim have made it onto all of his canvasses. Zola, Emile: Therse Raquin. Paris. 1979, 230-2. 31 Ginsborg, op. cit., ibid.

painters behavior but that it govern it normatively.32 Ginsborg concludes from this that, since design entails normativityand since purposiveness entails design[] the idea of normativity is already implicit in Kants initial notion of purposiveness [as defined in 10]33. Ginsborgs argument for the normative dimension of Kants definition of purposiveness in 10 thus rests flatly on the earlier ill-founded conclusion that Kant restricts purposiveness to intentional causality. As a result, Ginsborgs first argument for the connection between normativity and Kants notion of purposiveness fails. Yet, in spite of these shortcomings, something does seem undeniably right about the attempt to connect Kants notion of a causality of a concept with regard to its object to the idea of normativity. As it turns out, it is Ginsborgs tendency to frame this point in terms of intention and design that clouds the issue. We get a better sense of what Ginsborg may be on to by considering, again, her case of the obsessed painterprecisely because this is a case of a concepts unintended, sub-conscious influence on the production of its object, and not a case of design. Let us, moreover, follow Ginsborgs assumption that it is indeed the concept of the woman that influences the painter (as opposed to, say, the painters memorized image of the woman). Ginsborg contends that this sort of sub-conscious psychological influence does not involve normativity. This is, presumably, because she believes that for this influence to be a normatively governed process, [t]he concept has to function as a rule which represents to the painter how the painting ought to be.34 But this view of the production of the painting is altogether too agent-centric. For, even the non-intentional production of the womans likeness is subject to correctness conditions imposed by the concept of the woman35, provided that the concept does indeed influence the process, and provided that the product is indeed its (i.e., that concepts) object. The point is readily generalized. Any causal process on which a concept exerts a purported causal influence, and whose product is the concepts object36, is subject to correctness conditions imposed by that concept whether this process occurs in deliberate pursuit of an end or not. Normativity, indeed,
32 33

Ibid. Ibid. 34 Ibid., my emphasis. 35 Ibid. 36 On this condition, see 2.5.2, below.

seems to be built right into the notion of a concept that helps engender its object. Since Kant defines purposiveness as the causality of a concept with regard to its object, it appears that in Kants technical definition of purposiveness we get normativity for free. Thus begins the second, stronger line of Ginsborgs argument. Before investigating these matters further (2.5), it is high time we consider what Kant actually says about purposiveness in 10.

2.3 Kants Definition of Purposiveness 2.3.1 Concepts as Causes 10 of the Critique of the Power of Judgment is divided into two short paragraphs of four and five sentences, respectively. Kants definition of purposiveness is contained in the first three sentences of the first paragraph, which are, for that reason, the main focus here.37 Kant begins his account, following a prefatory remark about the transcendental nature of his definition (more on which in 2.3.2, below), by defining the notion of a purpose. Kant stipulates that he will call an object purpose whose concept is accorded an etiological role in the objects production:

[] a purpose is thus the object of a concept insofar as the latter [the concept] is regarded as the cause of the former [the object] (the real ground of its [the objects] possibility). (KU, AA 05: 220.1-3)38 The language of purposes thus comes into play when a concept is considered not merely in terms of its representational ties to its object but as a cause or real ground:

37

The fourth sentence of the first paragraph contains Kants definition of pleasure. The sentence is not relevant for an account of Kants technical conception of purposiveness since Kant insists, as we will see, that his notion of purposiveness abstract from [] anything empirical such as the feeling of pleasure [] (KU, AA 05: 219.32). Kants appeal to pleasure in the fourth sentence has, instead, to do with Kants attempt to tie his technical conception of purposiveness to his aesthetic theorya topic that need not concern us here. The second paragraph of 10 is also beyond the scope of an account of Kants technical definition of purposiveness (a point sometimes missed in the literature; see, e.g., Guyer, op. cit., 188; Ginsborg, op. cit., 332f.). In it, Kant addresses the explanatory versatility of that definition, and thus presupposes the definition as already understoodhe does not add to it. 38 [] so ist Zweck der Gegenstand eines Begriffs, sofern dieser als die Ursache von jenem (der reale Grund seiner Mglichkeit) angesehen wird [].

10

Where thus not merely the cognition of an object but the object itself (its form or its existence) as an effect is thought of as possible only through a concept of the latter, there one thinks of a purpose. (KU, AA 05: 220.4-7 )39 Kants name for this peculiar kind of causal relation between a concept and its object is purposiveness: [] and the causality of a concept with regard to its object is purposiveness []40. Thus, in the first two sentences of 10, Kant gives complementary definitions of purpose (as an object caused by its concept) and of purposiveness (as the causality of a concept with regard to its object).41 In the third sentence, Kant complicates matters by going beyond the simple bipartite (cause [concept]/effect [object]) structure of the causality of concepts that he has sketched thus far. Kant now claims that where a concept is considered as playing a causal role in the production of its object, a representation of the effect precedes the objects cause:

The representation of the effect is here the determining ground of its cause, and precedes the latter. (KU, AA 05: 220.8-9)42 Kant thus presents three separate relata as characteristic of conceptual causality, no longer just two: (1) a representation of the effect, (2) a cause of the effect, (3) the effect. It is initially unclear, whether this new account is supposed to be understood as a further definitionthis time of the notion causality of a concept with regard to its object (thereby effectively replacing the earlier bi-partite definiens in Kants definition of purposiveness with a new, tri-partite one)or a looser sort of characterization. More immediately troubling, depending on the interpretation we give to the three relata, at

39

Wo also nicht etwa blo die Erkenntni von einem Gegenstande, sondern der Gegenstand selbst (die Form oder Existenz desselben) als Wirkung nur als durch einen Begriff von der letztern mglich gedacht wird, da denkt man sich einen Zweck. 40 KU, AA 05: 220.3-4. [] und die Causalitt eines Begriffs in Ansehung seines Objects ist die Zweckmigkeit. 41 It is worth stressing that Kants characterization here is evidently simplified. A concept is, at best, a causal factor in the coming into being of its object. Kants simplification is, however, very much on purpose. As we will see, Kants aim in 10 is precisely to set aside other factors in order to highlight the conceptual factor. When, in the following, I speak of a concept causing its object, it should thus be borne in mind that this is shorthand, but in the spirit of Kants presentation. 42 Die Vorstellung der Wirkung ist hier der Bestimmungsgrund ihrer Ursache und geht vor der letztern vorher.

11

least two competing readings of Kants third sentence ensue. I sort out these two readings in the present section, and, in the next (2.3.2), consider what significance this pivotal sentence may have for Kants definitional project in 10. On perhaps the most straightforward reading of the third sentence of 10, Kant takes the representation of the effect to be the objects concept. Kants point, then, would be that where a concept is considered to be the cause of its object, that concept is also considered to be determining intermediate objects (i.e., [] its [the objects] cause[s]43) involved in the objects (i.e., [] the effect[s] 44) production. Thus, if I want to build a kitchen cabinet, my concept (kitchen cabinet) determines that I will need wood and woodworking tools to build it. My concept is then not only regarded as the real ground of any resulting cabinet but as a determining ground of the various objects causally involved in the cabinets production. Whatever its common-sense appeal, however, this reading lacks textual plausibility.45 If Kants point really concerned two types of cause and their logical interrelation (i.e., if Kants point were that an objects conceptual cause must also be considered the logical or determining ground of that objects physical cause), Kant could have hardly expected to make this point successfully by mentioning just one cause ([] its cause []46), and without qualification, at that. He could not have expected this, since the type of cause or real ground of an object that is ostensibly under consideration in the opening of 10 is the cause of an effect of the causality of conceptshence, a concept. In the context of a discussion of concepts as causes, and in the absence of any explicit consideration to the contrary, the only plausible referent for the expression its cause is an objects conceptual cause. Still, it is certainly not impossible that Kant, at the very heart of his short characterization of conceptual causes, equivocates on the notion under investigation in order to suggest that the cause of an effect of the causality of concepts is in fact a physical cause. It is, however, highly implausible that he would do this at the cost of turning the objects concept into a mere logical grounddevoid of

43 44

KU, AA 05: 220.8. Ibid. 45 Some support for it, however, can be drawn from Kants reference to purposiveness at KU, AA 05: 369.33-5. 46 KU, AA 05: 220.8.

12

causal efficacy altogetherat the very moment when his explicit aim is to highlight that concepts causality.47 If cause in Kants expression the determining ground of [the objects] cause thus names the objects concept, then the representation of the effect that (we are told) precedes and determines this cause must differ from the objects concept. A second and more plausible reading of the third sentence of 10 is, then, that by the representation of the effect that precedes and determines an objects conceptual cause, Kant means the guiding idea, goal or intention behind that objects production. What makes an intention the determining ground of an objects conceptual cause would be that it helps recruit the concept (the cause) whose object (the effect) has the potential (through its properties) of engendering the intended state of affairs. E.g., if the state of the world I intend to bring about is one in which I find storage for the china in my kitchen, the representation of this state (together with relevant background information) helps determine the concept kitchen cabinet as a concept the existence of whose object (a kitchen cabinet) has the potential of making the represented state real. On this reading, too, Kant would be guilty of an equivocation. The effect in Kants expression the representation of the effect here does double-duty as it refers both to the object that is to be produced and to the desired state that the objects existence is to make possible. But, unlike the former equivocation, which has Kant slide between the two fundamentally different types of cause (conceptual and physical) whose very distinction is at issue in the opening of 10, the latter equivocation is relatively benign. Kant would fail to distinguish between the existence of an object and the existence of a state of affairs, in a context where this distinction does not matter very much. I conclude that, given (1) the problems of the first reading, (2) the inherent textual plausibility and relatively minor problem of the second reading, as well as (3) the absence of plausible
47

A variation of this reading that avoids this problem is that by determining ground Kant does not mean a logical ground but, consistent with rationalist usage, a kind of cause. Continuing our assumption that by the representation of the effect Kant means the objects concept, that concept would then be a cause of its object (i.e., a real ground) precisely because it is, in the first instance, a cause of its objects physical cause (i.e., a determining ground). As a determining ground, the concept would then, however, be a cause of an object other than its own. The problem with this, as we will see in 2.5, below, is that the notion of a concepts causality in regard of an object other than its own no longer treats that concept as a cause qua representational entity. But where the representational nature of the concept is no longer at issue, the concept cannot be a determining ground. Thanks to Desmond Hogan for bringing this alternative reading to my attention.

13

alternatives to these interpretations, the second reading of the third sentence of 10 is the one we must adopt. The complication that Kant adds to the simple bipartite (cause [concept]/effect [object]) structure of the causality of concepts in the third sentence of 10 is, accordingly, that whenever a concept is considered causally efficacious with regard to its object, this causal efficacy is itself construed as embedded in the context of the intention an agent pursues in making the concepts object real. The production of an object through conceptual causality thus involves not only (i) a concept (as the real ground of its object) but (ii) an intention (as the determining ground of that concept). Note that, if we were to take this second, intentional component as definitional of the causality of a concept with regard to its object, Kant would be supplying just the sort of necessary condition for conceptual causality that Ginsborg was hoping to discover. Ginsborgs contention that Kants definition of purposiveness involves a normative dimension could, then, readily be established along her original lines, presented in 2.2, above. Accordingly, having established the meaning of the third sentence of 10, we must now consider its significance in order to see whether it supports Ginsborgs account. Specifically, we must determine whether Kants characterization of conceptual causality in the third sentence of 10 has the force of a definition.

2.3.2 Transcendental Determinations It would be a mistake to understand the third sentence of 10 as providing not only a sufficient but also a necessary condition for conceptual causality. While Kants sentence does suggest a strong connection between conceptual causality and intentionality, it is critical that we place the sentence in its proper context before drawing conclusions about the nature of the connection asserted. That context is provided by the very first (and most important) thing Kant says about his definition of purposiveness in 10, namely, that it explains [] what a purpose is []and what purposiveness is [] according to its transcendental determinations (without presupposing anything

14

empirical such as the feeling of pleasure) []48. Kant notes that one gives a transcendental definition of concepts [] when one has cause to suspect that they stand in relation to the pure faculty of cognition a priori []49. The suggestion in 10 is, then, that we need a transcendental definition of the concept of purposiveness because we have cause to suspect that this concept stands in relation to the new faculty of cognition a priori that Kant introduces in the third Critique: the reflecting power of judgment. To give this definition is to prepare the empirical concept of purposiveness for use in a transcendental context, and, hence, to strip it off all unduly empirical connotations. Kants characterization of his definitional project as transcendental thus reveals a marked dialectic between the first and third sentence of 10. In the first sentence, Kant insists on the transcendental nature of his definition of purposiveness in order to stress that the definition abstracts from contingent circumstances surrounding the causal efficacy of concepts. In the third sentence, Kant tells us what those contingent circumstances areforemost among them, the fact that the causal efficacy of concepts is encountered in the context of a wills intentional agency. Accordingly, the third sentence of 10, far from suggesting that Kants technical definition of purposiveness must be understood in terms of intentionality, suggests just the opposite. Given that Kants technical definition of purposiveness is concerned with the notions transcendental determinations, it deliberately abstracts from the intentional aspects surrounding conceptual causality, which (the third sentence reminds us) are its empirical determinations. On this interpretation, any attempt to derive normativity from the intentional dimension of conceptual causality that Kant mentions in the third sentence of 10 necessarily failsfor the simple reason that Kant mentions that dimension precisely in order to exclude it from his technical understanding of purposiveness. That this is indeed the dialectic of the opening of 10that by abstracting from [] anything empirical, such as the feeling of pleasure []50 Kant presents a definition of purposiveness that abstracts from the intentional dimension of the
48

KU, AA 05: 219.31-220.1. Wenn man, was ein Zweck sei, nach seinen transscendentalen Bestimmungen (ohne etwas Empirisches, dergleichen das Gefhl der Lust ist, vorauszusetzen) erklren will []. 49 EEKU, AA 20: 230.15-17. Es ist von Nutzen: zu Begriffen, welche man als empirische Principien braucht, eine transzendentale Definition zu versuchen, wenn man Ursache hat zu vermuthen, da sie mit dem reinen Erkenntnivermgen a priori in Verwandtschaft stehen. 50 KU, AA 05: 219.32.

15

production of artifacts altogetheris clear, given Kants account of the sensible determination of the will.51 For Kant, a sensibly (heteronomously) determined will desires a state of affairs in virtue of anticipated pleasurable effects.52 To satisfy the desire for this state of affairs (and reap the pleasurable effects) is the agents willi.e., her end or intention. Bringing about the object (or objects) that would make the state in question a reality is the agents means to that intended end. Attaining the means, in turn, is made possible through the adoption of a technical-practical rule53 that recommends an appropriate object: If you want state of affairs x, produce object y.54 The satisfaction of a desire to stay dry, for instance, is the intention behindhence, the rationale underlying the adoption of a technical-practical rule that recommendsbuilding a roof over ones head. When Kant explains that his technical definition of purposiveness considers purposiveness [] without presupposing anything empirical, such as the feeling of pleasure []55, he means that his definition abstracts from the anticipation of pleasure that constitutes the empirical context in which concepts become causally efficacious (by figuring in the consequents of technical practical rules). Accordingly, Kants nonempirical, transcendental definition abstracts from the desire in the service of which a concepts causal efficacy ordinarily stands. Since the satisfaction of a desire is, for Kant, the intention behind an objects production, it follows that Kants non-empirical, transcendental definition abstracts from the intention behind the objects production altogether.56 Kants transcendental definition of purposiveness is, then, carefully designed to exclude the intentional dimension of the causality of a concept with regard to its object. And it excludes more than that. By systematically abstracting from the end pursued in the production of an object, Kants transcendental conception of purposiveness excludes a whole range of conventional meanings of the notion of purposiveness. It excludes: (1) the
51 52

See, e.g., KpV, AA 05: 21-22. See KpV, AA 05: 22.16. 53 See KU, AA 05: 173.30. 54 Concepts that are considered causally efficacious with respect to their objects can then be understood, more precisely, as concepts that appear in the consequents of technical propositions adopted by sensibly determined wills. 55 KU, AA 05: 219.32-220.1. 56 See also KU, AA 05: 182.1-6.

16

properly teleological idea that a purposive object be directed at an end;57 (2) the functional idea that a purposive object be good or useful for an end; (3) the valuational idea that a purposive object be good at being good for an end; (4) the normative idea that, in light of its end, a purposive object be as it ought to be. Thus purged of all forward-looking concerns with an objects propensity as a means to the end of (desired) effects, Kants notion of purposiveness as the causality of a concept with regard to its object emerges as an unflinchingly backward-looking58 notionits sole concern is with an objects causal ancestry. To judge an object purposive, in this backward-looking sense, far from being a full-blooded judgment about its utility in an economy of ends, instead makes the oddly anodyne point that the objects concept played a causal role in the objects coming into being. Thus stripped off pedestrian connotations, Kants etiological conception of purposiveness is, emphatically, a critical philosophers term of arttailor-made for its use in Kants most recent a priori principlethe transcendental principle of purposiveness. Before considering whether Kant in fact succeeds in establishing an entirely backward-looking, non-teleological conception of purposivenessor whether residual normative life-blood animates the sparse logic of his technical vocabularya further question, posed forcefully by the idea of a backward-looking conception of purposiveness, must now be addressed: does the idea of a concept causing its object make any (Kantian) sense at all?

2.4 Mental Causation In discussing the idea of a causality of concepts or of a concepts causal role in the production of its object, Kant does not propose to provide a philosophical analysis of the suggested causal link between concepts and events in the physical world. Such an analysis would, on Kants view, be impossible.59 Very roughly, according to Kants Paralogisms in the Critique of Pure Reason, concept-involving psychological states occur in inner sense alone, and, therefore, cannot be located in space. But, according to
57 58

See n. 6, above. Thanks to Allison Simmons for helpful discussions of this terminology. 59 For a recent, forceful defense of this position, see Kenneth R. Westphal: Kants Transcendental Realism. Cambridge. 2004, 54n, 229ff.

17

the Analogies of the Critique of Pure Reason, determinate causal judgments can only concern both temporally and spatially identifiable substances interacting with one another. It follows that, on Kants view, we can make no determinate causal judgments abouthence cannot give an empirical account ofthe efficacy of concept-involving psychological states in the physical world. This is not to deny the possibility of deterministic mind-world causality. It is to deny that, were such determinism to be true, we could know or explain that fact. On theoretical grounds, Kant thus remains agnostic about the causal efficacy of concept-involving psychological states in the physical world. The concept of a causality of concepts is, accordingly, what Kant later calls a problematic60 concept or a concept for which [] no theoretical grounds can be given to determine whether an object corresponds to it []61. In spite of this peculiar theoretical status, the concept of a causality of concepts does play an important role in empirical cognition62: without the notion that our plans, schemes or concepts can be translated into realities, we would not be able to make sense of even the simplest cases of intentional human behavior. The notion of intentional agency itself would be meaningless.63 The idea that concept-involving psychological states have causal efficacy in the material world is, then, indispensable for our understanding of ourselves as agents and of that considerable part of reality we consider shaped by ourselves. This indispensability means that, however problematic the concept of a causality of concepts may be, it functions as what Kant, in the Teleology, calls a regulative concept64. A regulative concept, in the sense adopted there, is a concept that, although employed beyond the limits of its empirically legitimate use, nonetheless does considerable cognitive service, as it opens up areas of empirical research that would otherwise remain foreclosed. For instance, although considering organisms in functional terms goes well beyond the empirically legitimate use of our function vocabulary65,
60 61

See KU, AA 05: 360.21, 397.8. See Westphal, op. cit., 62n. 62 If Kants claims in the transcendental deduction of the principle of purposiveness are correct, it plays an even more fundamental transcendental role. 63 Conceptual causality is a necessary condition of intentional causality. 64 See KU, AA 05: 375.19. 65 Kant, of course, argues at great length in the Teleology that the concept of a natural purpose is a concept that we bring to organisms legitimately. But this legitimacy is not empirical since such judgments are a priori reflecting (teleological) judgments, and rest on the transcendental principle of purposiveness. The regulative use of the concept of a natural purpose (and in its wake the use of

18

function talk opens up areas of biological investigation that would be unavailable on a strictly mechanistic conception of the natural world. In the same way, while regarding physical effects as caused by concept-involving psychological states goes beyond the empirically legitimate use of the concept of cause, it does open up areas of cultural, sociological and every-day discourse that would otherwise be uninhabitable. By defining purposiveness in a backward-looking sense as the causality of a concept with regard to its object, Kant thus assigns the terminological top-spot in the Critique of the Power of Judgment to a concept that, at least in its ordinary employment (i.e., so far as concept-involving psychological states and events in the physical world are concerned), is problematic and of merely regulative use. The philosophically intriguing question this raises is, naturally, what Kant hopes to accomplish with this concept in the third Critique. For the purposes of this essay, however, I will not be concerned with the philosophical significance of Kants terminology.

2.5

The Causality of Concepts

We saw at the end of 2.2 that there is a perfectly straightforward sense in which normativity becomes relevant to Kants definition of purposiveness in 10. The fact that Kant defines purposiveness as the causality of a concept with regard to its object seems to build the idea of normativity right into his account. The notion of a concept causing its object certainly implies more than a merely causal link between concept and object: it implies a representational link between the two. Given this coupling of causal and representational dimensions, it would seem that Kants conception of purposiveness cannot be entirely backward-looking. Even if the notion abstracts from concerns with the ends pursued in conceptual causality, there would seem to remainat its corean appeal to a concepts irreducibly forward-looking determination of the specific sort of thing its

function vocabulary in biological discourse) is an extrapolation from this legitimate transcendental deployment in a priori reflecting (teleological) judgments (see KU, AA 05: 377.10-6). For Kant, the use of function vocabulary in our discourse on biological phenomena is, thus, either transcendentally necessitated or merely regulativeit does not rest on empirical grounds. One might think that function vocabulary is thus empirically legitimate only in the case of human artifacts. But the point of the present paragraph is that in the case of human artifacts, too, function vocabulary (insofar as it cannot be divorced from the idea of conceptual causes) ultimately lacks empirical legitimacy. Both in the case of artifacts, and in the derivative case of organisms, function vocabulary thus rests on an empirically unwarrantable application of the notion of cause.

19

object is supposed to be. If Kants technical conception of purposiveness did indeed entail an appeal to the normative constraint a concepts content places upon the object caused, then Ginsborgs conclusion, that [] the idea of normativity is already implicit in Kants initial notion of purposiveness [as defined in 10]66, would be unassailable.

2.5.1

Purposiveness as Conformity to a Conceptual Cause

Ginsborgs second, stronger argument explores this line of thought.67 She begins by distinguishing two separate aspects of the normative hold a concept may have on the coming into being of an object. An object may, in the first place, be subject to a concepts normative constraint. Ginsborg suggests that this is the case provided a concept plays a causal role in the objects coming into being.68 But she points out that it is a separate matter altogether whether the object thus caused satisfies the normative constraint in question. This is the case only if the object caused is, moreover, the one the causally efficacious concept represents. To support the idea that these two aspects of a concepts normative hold on the coming into being of an object really are distinct, Ginsborg gives the example of a potter who sets out to make a certain type of vase, but fails. Ginsborg explains that the resulting misshapen vase is an effect of the causality of a concept since the potter throws it [] with a concept in mind []69. She denies, however, that the causally efficacious concept in question represents the de facto outcome of the potters foundering efforts (i.e., she denies that the potters concept was the concept of a misshapen vase).70 But if the misshapen vase is not the effect of the concept misshapen vase (nor of any other possibly intervening concept, as was the case in Ginsborgs example of the love-struck painter)yet if it is an effect of the causality of a conceptthen it must be the effect of the potters concept of a shapely vase. There is, then, a discrepancy (all too familiar to the would-be artist) between the effect represented (a shapely vase) and the effect caused (a misshapen vase)the latter remaining painfully subject to the normative constraint of the former, even as it sadly fails to satisfy that constraint.
66 67

Ginsborg, op. cit., 342. See ibid., 343ff. 68 As we saw, Ginsborg also thinks that a designer must be conscious of that role; see 2.2, above. 69 Ibid., 345. 70 See ibid., 343.

20

With this distinction in place, Ginsborg turns to Kants definition of purposiveness as the causality of a concept with regard to its object. She points out that the case of the potter is not covered by Kants definition. The concept of a shapely vase that the potter employed, although causally implicated in generating the disappointing result, [] does not have causality in respect of its object since its object is not in fact produced []71. Since the careless ceramicists vase is subject to, but does not satisfy, the causally efficacious concepts normative constraint, the vase, although an effect of the causality of a concept, is not an effect of purposiveness as Kant defines it. Ginsborgs claim that the potters sputtering creative effort is not an instance of the causality of a concept with regard to its object (as opposed to, say, an instance of the causality of a concept with regard to its object that merely was impeded)but that it nevertheless is an instance of conceptual causalityentails that there are two ways in which a concept can be a cause: a concept can either cause (i) its object or (ii) objects other than its own. In other words, there are distinct conceptual causalities for creative success and for creative failure. If this interpretation were correct, then Kants aim in defining purposiveness in 10 would be to distinguish these two forms of conceptual causality from each other, and to reserve the name purposiveness for the former alone. Since, in the purposive sort of conceptual causality (but not in the non-purposive sort), the object caused is not only subject to but satisfies its causally efficacious concepts normative constraint, it follows that to judge an object purposive, in Kants sense, involves an irreducibly [] evaluative claim []72 that the object is as it is (by the standard of its conceptual cause) supposed to be. Ginsborgs conception of purposiveness as an evaluative notion is problematic, from the point of view of my interpretation, because, on her conception, a judgment of purposiveness makes substantive reference to the content of the concept judged to be causally efficacious. A judgment of purposiveness determines (i) what the concepts content tells us about how an object caused by it is supposed to be, and (ii) whether a given object purportedly caused by the concept really is that way. In the terminology adopted earlier, this substantive appeal to a causally efficacious concepts

71 72

Ibid., 345, my emphasis. Ibid., 344.

21

representational content is inherently forward-looking, as it looks toward objects (to be) effected by the concept as end-states. If judgments of purposiveness really did involve this appeal, then my contention that Kants notion of purposiveness is strictly backwardlooking, non-teleological and non-normative would have to be rejected.

2.5.2

The Nature of Conceptual Causality

But Ginsborgs argument is flawed as it moves from the straightforward idea that the causality of a concept with regard to its object may be thwarted by intervening causes to the unwarranted idea that a concepts causal efficacy may somehow cease to be causality of a concept with regard to its object altogether. Contrary to Ginsborgs suggestion, there is only one form of conceptual causality: the causality of a concept with regard to its object. The proper contrast to Kants causality of a concept with regard to its object is not another form of conceptual causality: it is non-conceptual causality, or mechanism. Once this is realized, it is easy to see that Kants technical account of purposiveness, in distinguishing conceptual from non-conceptual causality, entails no substantive, forward-looking appeal to a causally efficacious concepts representational content. Normativity, as we will see, still gets an honorary mention in Kants account of purposiveness, but it does not do any philosophical work. In order to appreciate this, let us consider the idea of a concept that causes an object other than its own in more detail. Ginsborgs example of the hapless potter has certain limitations (for one, compared with the time of kraters and amphorae, potters today take a less conceptual, more free-wheeling approach to the wheel). An example of rule-following in which an agent implements a well-defined concept or, better yet, a set of step-by-step instructions might thus serve us better. And since the point at issue is a certain type of failure of such implementationa type that so often stalks us in the kitchena culinary example will do well. Thus, I may follow a cake-recipe, and yet earn nothing but doughy rewards instead of the fluffy wholesomeness my pastry book promised. In such a case there is, then, a discrepancy between the effect represented and the effect caused of just the kind we are interested in. Do such discrepancies call for a more accommodating form of conceptual causality than Kant has on offerone that explains how concepts can cause objects they do not represent?

22

Consider the following three explanations of my cake calamity: (1) The recipe was flawed. Clearly, this explanation does not mean that the recipe I followed led to a result other than the one it represents. To the contrary, it means that there is a discrepancy between the recipe the author of my pastry book meant to put into print and the recipe that actually got printed, where the latter led to just the result it (however inadvertently) does representa doughy cake. (2) I was sloppy. On this scenario, the recipe is without fault and I am to blame (I may have been leaving out certain steps, or substituting others for unwitting inventions of my own, or both). Note, however, that this still does not mean that my pastry books recipe caused an effect other than the one it represents. Instead, my sloppiness effectively changed the recipe. Whether I was merely following the actual recipe with little precision or was acting as though I were following a different recipe with great precisionon neither construal is the resulting cake the unmediated effect of the steps given in my book. (3) The oven did not reach the temperature it indicated. On this scenario, the world is to blame, and yet the analysis of my mishap remains essentially the same as the previous one: failing to implement the recommendations of the recipe (even if through no fault of my own) either changes the recipe or, at any rate, thwarts the original recipes causal influence on the process. In dealing with cases of a discrepancy between effect represented and effect caused we thus consider as possible explanations either that the token of the concept that was employed contained a flaw, that the agent was incompetent, or that the world did not cooperate (or a combination of these). These explanations invoke intervening causes, issuing in (potentially severe) copying errors or translation mistakes, in order to explain the discrepancy at hand. We do not consider as a possible explanationbut would have to countenance it if Ginsborg were rightthat all in fact went well (i.e., that the token of the concept was the correct one, that the agent did as she was supposed to do, that the world cooperated), and that, in spite of all this, a different form of conceptual causality kicked in by which the concept helped generate an object other than its own. We do not dismiss this scenario as an explanation simply because, as a matter of contingent fact, such a thing does not happen. We dismiss it as an explanation of how a concept helps cause an object because we cannot conceive of this as a case of the causal efficacy of a concept qua concept. A regularity-analysis of causation teaches, of course, that random

23

and unprecedented consequents are always conceivable for a given causal antecedent. But a regularity-analysis does not have the same kind of purchase on this sort of case as it has on others. Pace Hume73, in the case of the causality of a concept the effect very much is contained in or prefigured bynamely, represented byits cause. An instance of the causality of a concept in which all goes well (i.e., concept-token, agent and world cooperate), and yet the concept helps bring into existence an object other than the one it representswhile not inconceivable on a regularity accountcannot conceivably be an instance of the causality of a concept qua concept or representational entity.74 I conclude that the causality of a concept just is the causality of a concept with regard to its object. To consider a concepts causal efficacy qua bearer of representational content is to consider the circumstances under which that content becomes reality. By contrast, the idea of a concept that mysteriously generates an effect it does not represent not only does not explain very much, it effectively treats the object as the effect of a non-representational and, therefore, non-conceptual form of causality. At this point, one may wonder, however, how my criticism of Ginsborgs claim that foundering design efforts must be excluded from the purview of purposive causality can possibly help support a non-teleological interpretation of Kants conception of purposiveness. My criticism of Ginsborgs claim is a rejection of her idea that objects may fail to satisfybut nevertheless remain subject toa causally efficacious concepts normative constraint. If conceptual causality indeed just is the causality of a concept with regard to its object, then the point at which an object truly fails to satisfy its causally efficacious concepts normative constraint (wherever precisely that point may be) must be the point at which it fails to be the effect of that concept (hence subject to its constraint) altogether. In denying that foundering design efforts are instances of purposiveness, Ginsborg thus mistakes cases in which an object fails to satisfy fully its

73 74

Hume, David: Enquiries Concerning Human Understanding. Oxford. 1975, 27-30. One may imagine a scenario in which my attempts to produce an F invariably produce a G, and in which, mastering this odd discrepancy, I set out to produce an F whenever I in fact want a G. Is the content of my concept of F then not directly and causally linked to the existence of an object other than its own? No. This is, after all, not a case in which all goes well. Indeed, we cannot describe the case without assuming some kind of mistranslation (however systematic) of F-instructions into G-instructions, hence, some kind of intervening non-conceptual cause that scrambles the original message and prevents its implementation. The very formulation of the scenario thus presupposes that a conceptual cause, qua concept, causes its object. Thanks, once again, to Des Hogan for raising the point.

24

causally efficacious concepts normative constraint for cases in which it fully fails to satisfy that constraint. She employs a maximalist conception of satisfaction where a minimalist conception would have been in order. The potters misshapen vase, after all, is still a vase: it accordingly does satisfy its causally efficacious concepts normative constraint (to a degree), and, therefore, is an instance of the causality of a concept with regard to its object. Yet, even a minimalist conception of satisfaction is a conception of satisfaction. My interpretation thus only seems to reinforce Ginsborgs most basic contentionthat Kants technical conception of purposiveness is irreducibly normative. On my view, as on hers, Kants conception of purposiveness entails that a purposive object satisfy its causally efficacious concepts normative constraint. How, then, can I claim that Kants conception is non-normative? There is an important difference between the way in which normativity enters Ginsborgs account, and the way in which it enters mine. On my reading, a judgment of purposiveness, in Kants technical sense, is concerned with an exceedingly fundamental ontological distinction. By pointing to its concepts role in an objects coming into being, a judgment of purposiveness makes the determination that the thing in question is a product of rational and not of mechanistic causalityof conceptually guided processes, not of nature churning aimless. A judgment of this sort proclaims the ontological status of the object to be that of an effect of mind and meaningit thus proclaims the objects artifacthood. This ontological judgment does entail a normative fact about the object in question: an effect of the causality of concepts is as it is (by the standard of its conceptual cause) supposed to be. But the normativity here rides on the metaphysicssatisfaction of the concepts normative constraint is cheap. Ex hypothesi, if the object did not satisfy the constraint (however minimally), it would not be a conceptual-content-become-real and, hence, an effect of the causality of a concept. Normativity thus enters the picture, as it were, by default. But then, whatever my determination of an objects artifacthood may be based upon75, it does not presuppose any substantive appeal to the normative constraint placed

75

As we know from the argument from design, circumstantial factors such as regularity, complexity and adaptation (let alone characteristics such as tool-marks and inscriptions) suffice to make the judgment. As David Hume reminds us at length, however, one must take great care here, lest the judgment turn out

25

on the object by its purported conceptual cause. It only presupposes that there be such a cause and, hence, such a constraint.76 Accordingly, it does not rest on an actual evaluation of whether the object is the way it is (by the standard of its conceptual cause) supposed to be (hence, of the degree to which the concepts constraint is satisfied). Rather, to judge that an object is purposive, in Kants transcendental sense, is to make the claim that the object has its origin in a concept (whatever concept that may be); hence, that it is the way its concepts content says it ought to be (whatever way that may be and however successful). Without thus requiring an evaluative appeal to the objects satisfaction of its causally efficacious concepts normative constraint, a judgment of purposiveness is strictly backward-looking, non-teleological, and non-normative. The nature of Kants non-teleological definition of purposiveness, then, is this. Kant acknowledges that it is in virtue of a causally efficacious concepts representational link to its effect that the causality of concepts must be treated as a special form of causality, distinct from mechanism. Having granted this much, however, Kant abstains from transforming this etiological point into a teleological one. Because it does not make substantive, evaluative appeal to the content of an objects purported conceptual cause, a judgment of purposiveness, in Kants sense, remains at the level of a strictly ontological, non-normative attribution of artifacthood. Kants transcendental definition of purposiveness thus successfully prevents teleology, which it was ostensibly designed to shut out by shutting out intention and design, from sneaking back in.

to be ill-founded. Hume, David: Dialogues Concerning Natural Religion. Richard H. Popkin (ed.). Indianapolis. 1998, 16-57. 76 We can see this in Kants discussion of ancient stone utensils of unknown purpose (see KU, AA 05: 236.11n), and of an apparently authorless hexagon drawn in the sand (see KU, AA 05: 370.16-32).

26