journal of the philosophy of history 9 (�0�5) 37�–39� brill.com/jph One Fell Swoop Small Red Book Historicism Before and After Davidson Constantine Sandis University of Hertfordshire c.sandis@herts.ac.uk Abstract In this essay I revisit some anti-causalist arguments relating to reason-giving explanations of action put forth by numerous philosophers writing in the late ‘50s and early ‘60s in what Donald Davidson dismissively described as a ‘neo-Wittgensteinian current of small red books’. While chiefly remembered for subscribing to what has come to be called the ‘logical connection’ argument, the positions defended across these volumes are in fact as diverse as they are subtle, united largely by a an anti-scientistic spirit which may reasonably be described as historicist. I argue that while Davidson’s causalist attack was motivated by an important explanatory insight borrowed from Hempel, it caused serious damage to the philosophy of action by effectively brushing over a number of vital distinctions made in the aforementioned works. In seeking to revive these I propose an approach to the theory of action explanation that rescues the anticausalist baby from the historicist bathwater. Keywords historicism – reasons – action – Davidson – Wittgensteinian – anti-causalism This paper was presented in 2013 at the Society for the Philosophy of History APA Session in San Francisco, (March 17–23), the Welsh Philosophical Society Colloquium, Gregynog Hall, (3–5 May), and the ‘50 years of Davidson’s “Action, Reason, Causes” ’ conference in DuisbergEssen University (August). Many thanks to all the organisers and participants, in particular Mark Bevir, David Cockburn, Giuseppina D’Oro, Hans-Johann Glock, Hans Sluga, Ralf Stoecker, Thomas Spitzley, and Karsten Stueber. I’m also very grateful to two anonymous referees for perceptive criticisms of an earlier draft. © koninklijke brill nv, leiden, ���5 | doi �0.��63/�87��636-��34�308 JPH_009_03_F4-Sandis_372-392.indd 372 10/13/2015 12:31:41 PM One Fell Swoop 373 Prologue Donald Davidson begins his landmark paper ‘Action, Reasons and Causes’ by contrasting his so-called causalist account of reason-giving explanations of action to the views of a large number of then prominent philosophers, all writing from within a loosely Wittgensteinian tradition. These are dismissed in one fell swoop: In this paper I want to defend the ancient – and commonsense – position that rationalization is a species of causal explanation. The defence no doubt requires some redeployment, but it does not seem necessary to abandon the position, as has been urged by many recent writers [fn: Some examples: Gilbert Ryle, The Concept of Mind, G.E.M. Anscombe, Intention, Stuart Hampshire, Thought and Action, H.L.A. Hart and A.M. Honoré, Causation in the Law, William Dray, Laws and Explanation in History, and most of the books in the series edited by R.F. Holland, Studies in Philosophical Psychology, including Anthony Kenny, Action, Emotion and Will, and A.I. Melden, Free Action].1 In a later essay, Davidson acknowledges Carl Hempel’s deep influence on his views whilst re-affirming the contrast to the volumes in Holland’s series: In December of 1961 Hempel gave the presidential address at the annual meeting of the Eastern Division of the American Philosophical 1 D. Davidson, “Actions, Reasons, and Causes”, Journal of Philosophy 60: 685–700; as reprinted in his Essays on Actions and Events, 2nd revised edition (Oxford: Clarendon Press, 2001), 3. The full Holland series, published by Routledge & Kegan Paul, ran as follows: 1. D.W. Hamlyn, The Psychology of Perception (1957); 2. P. Geach, Mental Acts (1957); 3. R.S. Peters, The Concept of Motivation (1958); 4. A. MacIntyre, The Unconscious (1958); 5. P. Winch, The Idea of a Social Science (1958); 6. N. Malcolm, Dreaming (1959); 7. A.I. Melden, Free Action (1961); 8. D.M. Armstrong, Bodily Sensations (1962); 9. A. Kenny, Action, Emotion and Will (1963); 10. P. Alexander, Sensationalism and Scientific Explanation (1963); 11. J. Bennett, Rationality (1964); 12. J. Kovessi, Moral Notions (1967); 13. H. Fingarette, Self-Deception (1969); 14. T. Penelhum, Survival and Disembodied Existence (1970); 15. I. Dilman & D.Z. Phillips, Sense and Delusion (1971); 16. M. O’C Drury, The Danger of Words (1973); 17. J. Teichman, The Mind and the Soul (1974); 18. C.V. Wilkes, Physicalism (1978); 19. R. Bambrough, Moral Scepticism and Moral Knowledge (1979); 20. R.W. Newell, Objectivity, Empiricism and Truth (1986). From here onwards any general reference to small red books is intended to apply to nine of the first ten (published by 1963), the exception being Armstrong whose sympathies were far removed from those of the others. journal of the philosophy of history 9 (2015) 372–392 JPH_009_03_F4-Sandis_372-392.indd 373 10/13/2015 12:31:41 PM 374 Sandis Association. The title was ‘Rational Action’. In that address, Hempel argued that explanation of intentional action by appeal to the agent’s reasons does not differ in its general logical character from explanation generally; in taking this position, he was swimming against a very strong neo-Wittgensteinian current of small red books.2 The contrast is not imaginary, though we shall come to see in due course that it is subtler than Davidson lets on. Numerous books in Holland’s series defended the view that the relation between an agent’s reason(s) and her action(s) should not be understood causally but logically, normatively, conceptually, and/or hermeneutically.3 In this they are united by various forms of what might reasonably be termed anti-scientism. The books don’t have a clearly articulated and specific target here – certainly no definition of scientism is ever provided – but there is a general suspicion of any philosophy that attempts to answer questions relating to human minds and actions by appeal to models stemming from natural science e.g. physical mechanisms.4 Here is how one reviewer at the time characterised this movement: The entire series adds up to nothing less than a major philosophical revolution set off by Ryle’s Concept of Mind and Wittgenstein’s Investigations. No doubt devotees of natural science will regard this development as more of a counter-revolution, since modern philosophy began with the liquidation of purposive concepts from natural science while the goal of this new movement is to expunge pseudo-mechanistic concepts from the language of human affairs. The illuminating discoveries made by the writers of these monographs should have a profound effect on psychology and the social sciences, in revealing the absurdity of all attempts to model them on the natural sciences.5 In what follows, I shall outline the sense in which this anti-scientistic outlook may be plausibly described as a historicist one. I next argue that despite this 2 D. Davidson, “Hempel on Explaining Action”, Erkenntnis 10 (1976); as reprinted in his Essays on Actions and Events, 261; the expression he meant to use is actually ‘little red books’. For a proto-Davidsonian approach to Wittgenstein’s influence on the small red books see W. Cerf, “Studies in Philosophical Psychology”, Philosophy and Phenomenological Research, Vol. 22, No. 4 (June 1962), 558. 3 See § II below. 4 Some of the authors controversially assume that any appeal to causation counts as evidence for this. I return to this issue in sections 3 and 4. 5 R. Abelson, “Free Action by A.I. Melden”, Philosophy and Phenomenological Research, Vol. 23, No. 4 (June 1963), 616. journal of the philosophy of history 9 (2015) 372–392 JPH_009_03_F4-Sandis_372-392.indd 374 10/13/2015 12:31:41 PM One Fell Swoop 375 unity and a related allegiance to what has come to be known as the ‘logical connection argument’, there remain some sharp disagreements to be found across the books which Davidson so readily grouped together. These differences are frequently underscored by numerous insightful distinctions which have been largely ignored by post-Davidsonian philosophers of action, to the detriment of the subject. In defending a selection of these, I shall eventually suggest an approach to the explanation of action which steers clear of the Scylla of scientism without crashing into the Charybdis of the historicism in question. I Historicist Blends In the 1950s and ‘60s, much of the debate concerning action explanation was conducted within the philosophies of history and social science.6 The term ‘historicism’ has confusingly been used to describe both the view that the explanation of historical action appeals to laws of history (which may or may not be distinct from natural laws),7 and the diametrically opposed view that – unlike explanation in the natural sciences – historical explanation is not law-like at all.8 It is only in this latter sense that I shall be using the term here. So understood, historicism has been further defined as the view that ‘historiography is characterized as a matter of understanding or interpreting events . . . treated as distinct from the aim of explanation or prediction within the natural sciences’.9 6 For details and wider context see G. D’Oro & C. Sandis, “From Anti-Causalism to Causalism and Back: A Century of the Reasons/Causes Debate”, in (eds.) D’Oro & Sandis Reasons and Causes: Causalism and Non-Causalism in the Philosophy of Action (Basingstoke: Palgrave Macmillan, 2013), esp. 17–21. 7 Karl Popper dismissively uses it in this sense in both The Open Society and its Enemies, Centenary Edition (London: Routledge, 1945/2002), Ch.1, and The Poverty of Historicism (London: Routledge & Kegan Paul, 1957). 8 As if this weren’t confusing enough the term has also been used to characterise ‘any position which promotes historical thinking in philosophy and warns against ignoring or distorting the past’ (H-J. Glock, “Wittgenstein and History”, in (eds.) A. Pichler & S. Saatela, Wittgenstein: The Philosopher and his Works (Bergen: University of Bergen Press, 2005, 238. I shall not be concerned with this altogether different debate here though its complicated relation to Wittgenstein and those whom he inspired is not without interest (and is hinted at in the first quotation by Cerf further below). 9 R. D’amico, “Historicism”, in (ed.) A. Tucker, A Companion to the Philosophy of History and Historiography (Oxford: Wiley-Blackwell, 2011), 243; cf. M. Bevir, “Anglophone Historicism: From Modernist Method to Post-analytic Philosophy”, Journal of the Philosophy of History 3 (2009), 211–224. journal of the philosophy of history 9 (2015) 372–392 JPH_009_03_F4-Sandis_372-392.indd 375 10/13/2015 12:31:41 PM 376 Sandis Defenders of this mode of historicism have included Oakeshott, Croce, Collingwood, and Dilthey.10 In the words of G.H. von Wright, their common protest ‘is directed against the application to history of a category which is at home in the natural sciences’.11 This outlook – in unexpected alliance with both Wittgenstein and eighteenth-century Christianity – figures as a strong influence on many of the red book authors, though not all would accept the points about either prediction or understanding as outlined above.12 In a characteristically cynical review of the first six volumes of the series, the Princeton phenomenologist Walter Cerf captured this strange inheritance as follows: . . . the Studies in Philosophical Psychology are not, to be sure, studies in rational psychology. They contain no explicit commitment to metaphysics, in whatever sense of the term, no commitment, explicit or implicit, to a priori knowledge and certainly no proofs of the substantiality and immortality of the soul . . . [but] The Christian tenor of eighteenthcentury rational psychology is still audible in the antiscientistic (but not antiscientific) message of the Studies: Beware the encroachments of the natural sciences upon the specifically human. Winch presents us with a veritable declaration of independence of the social sciences from the natural . . . All this sounds rather like Dilthey (even though, quite typically, 10 11 12 M. Oakeshott, Experience and its Modes (Cambridge: Cambridge University Press, 1933), B. Croce, History as the Story of Liberty, trans. S. Sprigge (London: Allen & Unwin, 1938/1941); W. Dilthey, Patterns and Meaning in History: Thoughts on History and Society, ed. H.P. Rickman (London: Harper Row, 1961). The leading contemporary defender of this variety of historicism is Giuseppina D’Oro. See, for example, her “Two Dogmas of Contemporary Philosophy of Action”, Journal of the Philosophy of History (2007), vol. 1, 11–26, “Davidson and the Autonomy of Human Sciences”, in (ed.) J. Malpas, Dialogues with Davidson: New Perspectives on his Philosophy (Boston MA: MIT Press, 2011), and “Reasons and Causes: The Philosophical Battle and the Meta-Philosophical War”, Australasian Journal of Philosophy (2012). G.H. Von Wright, Explanation and Understanding (London: Routledge & Kegan Paul, 1971), 201. Von Wright’s own view of the debate about causation in history is that a significant part of is due to conceptual and terminological muddles. Ironically, the very same diagnosis would later be reached by his student Fred Stoutland on the debate between von Wright and Davidson: F. Stoutland, “Intentionalists and Davidson on Rational Explanation”, in (ed.) G. Meggle, Actions, Norms, Values (Berlin: Walter de Gruyter, 1999), 191–208. The relation of both tenets to the history of philosophy of action is detailed in D’Oro & Sandis, “From Anti-Causalism to Causalism and Back: A Century of the Reasons/Causes Debate”. journal of the philosophy of history 9 (2015) 372–392 JPH_009_03_F4-Sandis_372-392.indd 376 10/13/2015 12:31:41 PM 377 One Fell Swoop none of the six seems to be acquainted with him), who tried to separate a nonscientistic psychology, ‘understanding’ or ‘hermeneutic’ psychology, from explanatory psychology. However, Dilthey did not, I believe, wish to restrict explanatory psychology in any way, while some of our authors out-diltheying Dilthey, go so far as to contend that no explanatory theory of the natural science variety has any logical business in the domain of the specifically human . . . man is set apart from nature. Behind their antiscientistic attitude towards psychology is an antinaturalistic conception of the nature of man . . . the belief that the methods of the natural sciences are in principle inadequate to an understanding of what is specifically and essentially human.13 Dilthey’s own statement of his rather technical distinction between explanation (‘eklaren’) and understanding (‘verstehen’) is axiomatic in its brevity: We explain nature, but understand the life of the soul (Dilthey 1961, vol. 5, 144). On Cerf’s understanding of this outlook, human history and psychology may be treated as both natural sciences (NS) which are the province of event explanations (E) and human sciences (HS) concerned with our understanding (U) of people: Psych/Hist E U NS HS To understand a person, in the sense in question, is to render her intelligible, for example via some narrative which helps us make better sense of her. This is not to offer a causal explanation of anything. Conversely, one may provide a causal explanation of an action understood as a natural event without contributing to our understanding the mind or soul of the agent behind it.14 13 14 Cerf, “Studies in Philosophical Psychology, 537–8 & 556. It should be noted that both Dilthey’s outlook and that of the small red books is not limited to the explanation of human action, for example Hamlyn (1957) applies it to his approach to perception, and Malcolm (1959) to dreaming. I restrict the discussion merely so as to stay on topic. journal of the philosophy of history 9 (2015) 372–392 JPH_009_03_F4-Sandis_372-392.indd 377 10/13/2015 12:31:41 PM 378 Sandis Cerf maintains that the red book authors go much further than this in distinguishing between two kinds of explanation, the natural (NE) and the human (HE). In so doing, they reserve the term understanding (U) for the latter. Crucially, this domain comprises the entire domain of history and psychology. These are deemed to not be amenable to scientific explanations, whose nature are by definition causal (C): Explanation NE HE Science (C) Psych/Hist (U) This account of the relation of small red book historicism to Dilthey is by and large a plausible one. Nevertheless, authors differ in their precise accounts and it is difficult to ascertain how much of the disparity is ultimately terminological. Which of the resulting views may be dubbed as anti-naturalistic, for example, seems to be a moot point. This should not surprise us given the variety of things that we might contrast ‘natural’ with. As Jennifer Hornsby writes: . . . we need to consider carefully this thought that persons and their actions are part of nature. It seems right when we point out that nothing supernatural needed to have happened for human beings to evolve, and that it is a natural fact about people that, for instance, they have the abilities they do, and thus a natural fact that there are actions. Such considerations ensure that a naturalistic view of ourselves is in order . . . but they do not help to place our actions in a world ‘of nature’ if a world ‘of nature’ is to be thought of as constituted independently of the human beings that occupy it.15 The small red book use of Dilthey is sometimes (e.g. in the case of Winch) filtered through a related debt to Collingwood (and by extension Hegel). On Collingwood’s view (which may be traced back to Hegel16) all actions have an inner (psychological) and an outer (behavioural) side, only the second 15 16 J. Hornsby, Simple Mindedness: In Defense of Naive Naturalism in the Philosophy of Mind (Cambridge MA: Harvard University Press, 1997), 148–9; cf. J. McDowell, Mind and World, 2nd Edition (Cambridge MA: Harvard University Press, 1996, 73–77. See A. Laitinen & C. Sandis (eds.), Hegel on Action (Palgrave Macmillan, 2010). journal of the philosophy of history 9 (2015) 372–392 JPH_009_03_F4-Sandis_372-392.indd 378 10/13/2015 12:31:41 PM One Fell Swoop 379 of which is amenable to study by the natural sciences. The two sides are not related causally in any standard empirical sense. Rather, as independently also hinted at by Wittgenstein,17 the latter is an expression of the former: . . . an action is the unity of the outside and inside of an event . . . To the scientist, nature is always and merely a ‘phenomenon,’ not in the sense of being defective in reality, but in the sense of being a spectacle presented to his intelligent observation; whereas the events of history are never mere phenomena, never mere spectacles for contemplation, but things which the historian looks, not at, but through, to discern the thought within them. In thus penetrating to the inside of events and detecting the thought which they express, the historian is doing something which the scientist need not and cannot do . . . This does not mean that words like ‘cause’ are necessarily out of place in reference to history; it only means that they are used in a special sense . . . When an historian asks ‘Why did Brutus stab Caesar?’ he means ‘What did Brutus think, which made him decide to stab Caesar?’ The cause of the event, for him, means the thought in the mind of the person by whose agency the event came about: and this is not something other than the event, it is the inside of the event itself.18 Though receptive to much of this, Winch is worried by the role in providing understanding which he takes Collingwood to ascribe to re-enactment. I quote at some length in order to do justice to the nuances of his position: Collingwood’s conception of all human history as the history of thought . . . is no doubt an exaggeration and the notion that the task of the historian is to re-think the thoughts of the historical participants is to some extent an intellectual distortion . . . But Collingwood is right if he is taken to mean that the way to understand events in human history . . . is more closely analogous to the way in which we understand expressions of ideas than it is to the way we understand physical processes . . . Collingwood . . . says that the aim of the historian is to think the very same thoughts as were once thought . . . But though extinct ways of thinking may, in a sense, be 17 18 L. Wittgenstein, Philosophical Investigations, 4th edition, tr. G.E.M. Anscombe, P.M.S. Hacker, & J. Schulte (Oxford: Wiley-Blackwell, 1953/2009), §§ 335ff. & Zettel (Oxford: Basil Blackwell, 1967), §53–60. R.G. Collingwood, The Idea of History, rev. edition (Oxford: Clarendon Press, 1946/1994), 213–4; cf. Von Wright, Explanation and Understanding, 87–8. journal of the philosophy of history 9 (2015) 372–392 JPH_009_03_F4-Sandis_372-392.indd 379 10/13/2015 12:31:41 PM 380 Sandis recaptured by the historian, the way in which the historian thinks them will be coloured by the fact that she has had to employ historiographical methods to recapture them. The medieval knight did not have to use those methods in order to view his lady in terms of courtly love: he just thought of her in those terms . . . And naturally, it is even more impossible for me to think of his lady as he did. Nevertheless, Collingwood’s view is nearer the truth than is that most favoured in empiricist methodologies of the social sciences, which runs somewhat as follows . . . the historian unearths . . . data and presents them to his more theoretically minded colleagues who then produce scientific generalizations and theories establishing connections between one kind of social situation and another . . . social relations must be an unsuitable subject for generalizations and theories of the scientific sort to be formulated about them: Historical explanation is not the application of generalizations and theories to particular instances: it is the tracing of internal relations.19 I shall return in the next section to Winch’s more positive argument, but it is worth stopping to observe here that his interpretation of Collingwood is uncharitable. This is because the view in question is not as metaphysically mysterious as Winch supposes. In Collingwood’s own words: . . . for the historian there is no difference between discovering what happened and discovering why it happened . . . . . . how does he historian discern the thoughts which he is trying to discover? . . . by re-thinking them in his own mind . . . This re-enactment . . . is not a passive surrender to the spell of another’s mind; it is a labour of active and therefore critical thinking. The historian not only re-enacts past thought, he re-enacts it in the context of his own knowledge and therefore, in re-enacting it, criticizes it, forms his own judgment of its value, corrects whatever errors he can discern in it.20 This view comes remarkably close to that later found in Anscombe: 19 20 Winch, The Idea of a Social Science, 123–4. He has more positive things to say about Collingwood in his “Understanding a Primitive Society”, American Philosophical Quarterly, Vol. 1 (1964); as reprinted in his Ethics and Action (London: Routledge & Kegan Paul, 1972), 14. Collingwood (1946/1994: 177 & 215). The view is taken further by W.H. Dray in his Laws and Explanation in History (Oxford: Clarendon Press, 1957), 108ff. journal of the philosophy of history 9 (2015) 372–392 JPH_009_03_F4-Sandis_372-392.indd 380 10/13/2015 12:31:42 PM One Fell Swoop 381 The description of something as a human action could not occur prior to the existence of the question ‘Why?’, simply as a kind of utterance by which we were then obscurely prompted to address the question.21 With Croce, Dilthey, and Collingwood also came the idea that historical understanding requires empathy of some kind. Thus, for example, Isaiah Berlin, would later write of Vico and others that they were guided by: The very notion that the task of the historian was not merely to establish facts and give causal explanations for them, but to examine what a situation meant to those involved in it, what their outlook was, by what rules they were guided . . . not merely knowledge of facts and events, but understanding – Einfühlung, empathy – is required.22 Whether or not understanding requires empathy has since been a matter of considerable debate.23 Whatever the right answer may be, the small red book emphasis on understanding has been vindicated to the extent that empathy has in recent years become a major topic of philosophical research. The victory, however is arguably a Pyrrhic one, for the standard approach to empathy remains stubbornly reductive: empathy is a tool we use to gain access to the informational content of other peoples’ mental states (which double as their motivating reasons).24 21 22 23 24 G.E.M. Anscombe, Intention (Oxford: Basil Blackwell, 1957), § 46, p.83 (emphasis in original). For the view that Collingwood’s views here are also in unexpected alliance not only with Hegel, Dray, C. Taylor, Wittgenstein and Ryle, see C. Sandis , ‘A Just Medium: Empathy and Detachment in Historical Understanding’, Journal of the Philosophy of History, Volume 5, Issue 2, 2011, 179–200. I. Berlin, “The Divorce between the Sciences and the Humanities”, Tykociner lecture published by the University of Illinois (1974); as reprinted in The Proper Study of Mankind, eds. H. Hardy & R. Hauscheer (London: Pimlico, 1988), 355–6. See M. Bevir, The Logic of the History of Ideas (Cambridge: Cambridge University Press, 1999), K.R. Stueber, Rediscovering Empathy: Agency, Folk psychology, and the Human Sciences (Cambridge MA: MIT, 2006), and Sandis , ‘A Just Medium’. This view is, for example, endorsed in many of the essays in (eds.) A. Coplan & P. Goldie, Empathy (Oxford: Oxford University Press, 2011). In their introduction, Coplan and Goldie explicitly relate the empathy research agenda to various movements in the history of ideas, including historicism (under the label ‘hermeneutic tradition’) but, as is the case with all the essays in their volume, omit to mention any of the small red books. This omission is a sign of the how widespread the collateral damage of Davidson’s attack has been. journal of the philosophy of history 9 (2015) 372–392 JPH_009_03_F4-Sandis_372-392.indd 381 10/13/2015 12:31:42 PM 382 Sandis My aim in this section has been to give a taste of the sense in which many of the small red books may be said to be historicist. It is now time to focus on some of the particular claims about reasons for action that Davidson objected to. Whilst situated within the framework outlined above, I hope to demonstrate that they ultimately hang on additional considerations. For reasons of space I shan’t be able to properly rehearse the arguments of each of the authors here, so I shall hint at some of the diversity while focusing on those I find most insightful. I begin with a cluster of moves collectively known as the ‘logical connection’ argument. This centrepiece of anti-causalism was understandably a chief target of Davidson’s. I shall be maintaining that while Davidson was right to criticize the negative argument at the centre of this cluster, he ignored a positive argument which, albeit more peripheral, contains the seeds of a fruitful approach to the relation between actions and the reasons for which we perform them. In so doing, he threw the anti-causalist baby out with the historicist bathwater. II The Logical Connection The historicist framework of the small red books lends itself to the view that actions are not mere pieces of bodily movement to me explained according to mechanistic laws of nature. Whatever one makes of this outlook it is important to remember that the anti-causalism that Davidson was combatting was not about action explanation in general but about explanation in terms of the agent’s reasons for acting as she did (from here onwards ‘agential reasons’): . . . some philosophers have concluded that the concept of cause that applies elsewhere cannot apply to the relation between reasons and actions, and that the pattern of justification provides, in the case of reasons, the required explanation.25 Indeed, some of the philosophers Davidson has in mind here (e.g. R.S. Peters), deny that agential reasons are causes, while allowing for causal explanations of human behaviour that appeal to motives and other factors which they distinguish from agential reasons. Moreover, even Davidson was later careful to state that the mantra ‘reasons are causes’ should not be understood as a strict metaphysical statement about causal relations but, rather, as a point about the necessarily causal aspect of the explanatory role of reasons. Unlike 25 Davidson, “Actions, Reasons, and Causes”, 9. journal of the philosophy of history 9 (2015) 372–392 JPH_009_03_F4-Sandis_372-392.indd 382 10/13/2015 12:31:42 PM One Fell Swoop 383 causal relations, causal explanations are sensitive to description.26 Davidson’s account is arguably anti-scientistic in maintaining, pace Hempel, that there are no strict psycho-physical laws and that the explanation of human action requires no knowledge of purely physical ones. This understanding of the relation between psychological and physical aspects of the world (viz. anomalous monism) in some respects resembles the views of Hegel, Collingwood, and Dilthey.27 Be that as it may, Davidson’s insistence that the explanations in question only work if backed by strict physical laws (which we may be ignorant of) is enough to distance him from the mode of historicism defined at the outset of § I. An important bone of contention here, between Davidson and the red book authors, is the latters’ appeal to the so-called ‘logical connection argument’, according to which the relation between reasons and causes is not causal but internal, logical, conceptual, intrinsic, hermeneutic, and/or normative.28 The most famous proponent of this anti-causalist stance was Melden. His influential version of the argument connects the logical connection point with the aforementioned division of action into inner and outer parts related by expression, not causation: . . . the very notion of a causal sequence logically implies that cause and effect are intelligible without any logical internal relation of the one to the other . . . The interior event which we call ‘the act of volition’ . . . must be logically distinct from the alleged effect – this is surely one of the lessons we can derive from a reading of Hume’s discussion of causation. Yet nothing can be an act of volition that is not logically connected with 26 27 28 D. Davidson, “Causal Relations”, Journal of Philosophy, 64 (1967); as reprinted in his Essays on Actions and Events, 149–62. Brian McLaughlin argues that while explanations in terms of primary reasons are causal, Davidson is wrong to infer from this that rationalization is a species of causal explanation (B.P. McLaughlin, “Why Rationalization is Not a Species of Causal Explanation”, D’Oro & Sandis, Reasons & Causes, 97–123). For a small red book take on the relation between explanation and description in science see Alexander (1963: Ch. VI). For two distinct ways in which Davidson’s account of action and its explanation is arguably not far removed from that of Hegel see M. Quante, Hegel’s Concept of Action, trns. D. Moyer (Cambridge: Cambridge University Press, 2004), and D. Knowles, (2010), “Hegel on Action, Reasons, and Causes”, in (eds.), A. Laitinen & C. Sandis, Hegel on Action (Basingstoke: Palgrave Macmillan, 2010), 42–58. Unhelpfully, many of these different attributes are frequently run together across the red books. journal of the philosophy of history 9 (2015) 372–392 JPH_009_03_F4-Sandis_372-392.indd 383 10/13/2015 12:31:42 PM 384 Sandis what is willed – the act of willing in intelligible only as the act of willing whatever it is that is willed.29 In a larger red book published in a different Routledge series,30 von Wright points out that while this point is frequently made by reference to willing or intention, it ought to apply equally well to decisions, desires, motives, reasons, and so on.31 Indeed, Davidson rehearses Melden’s argument as follows: Since a reason makes an action intelligible by redescribing it, we do not have two events, but only one under different descriptions. Causal relations, by contrast, demand distinct events.32 It is worth remembering here that at the time of writing ‘Actions, Reasons, Causes’ Davidson was in agreement with Anscombe’s claim that the intention with which an action is performed ‘does not refer to an entity or state of any kind’.33 His objection to the logical connection argument is not that no such connection exists but that it doesn’t preclude a causal one. His line of argument hangs on his aforementioned distinction between causal relations and causal explanations (the latter, but not the former, being sensitive to description): To describe an event in terms of its cause is not to confuse the event with its cause, nor does explanation by redescription exclude causal explanation.34 To this he adds that ‘[i]n any case there is something very odd in the idea that causal relations are empirical rather than logical . . . The truth of a causal 29 30 31 32 33 34 Melden, Free Action, 52–3. The argument is inspired by Wittgenstein’s Philosophical Investigations and G. Ryle, The Concept of Mind (London: Hutchinson, 1949), 106–11; variations of it may be found in Anscombe, Intention, §§ 19 & 29, Winch, The Idea of a Social Science, 124, R.S. Peters, The Concept of Motivation, 27ff., S, Hampshire, Thought and Action (London: Chatto & Windus, 1959), 166ff., and Kenny, Action, Emotion and Will, 76ff. A later versions of this argument may be found in A.R. White, The Philosophy of Mind (New York A.A. Knopf, 1967). See D’Oro & Sandis, D’Oro & C. Sandis, “From Anti-Causalism to Causalism and Back: A Century of the Reasons/Causes Debate”, 35, n.46. Von Wright, Explanation and Understanding, 95. Davidson, “Actions, Reasons, and Causes”, 13–14. Davidson, Essays on Actions & Events, xvii. He came to abandon this view in his later paper “Intending”, in (ed.) Y. Yovel, Philosophy of History and Action (D. Reidly and The Magnes Press, 1978); as reprinted in his Essays on Actions and Events, 83–102. Davidson, “Actions, Reasons, and Causes”, 14. journal of the philosophy of history 9 (2015) 372–392 JPH_009_03_F4-Sandis_372-392.indd 384 10/13/2015 12:31:42 PM One Fell Swoop 385 statement depends on what events are described.’35 As von Wright’s student Fred Stoutland would later claim, ‘the mere fact of a logical connection doesn’t rule out ‘a contingent relation between intentions and the occurrence of what fulfills them’.36 Stoutland influenced his teacher, who concluded that while ‘those who advocate what has become known as the Logical Connection Argument are substantially right’ it is not obvious ‘whether anybody has yet succeeded in presenting the argument quite convincingly’ and ‘[s]ome versions of it are not only unconvincing but manifestly defective’.37 So, is there anything to be said for the logical connection argument? Its negative claim that a logical connection cannot also be causal is, at best, difficult to establish. But the positive claim that in citing reasons we explain action by pointing to non-causal facts or relations is more promising. An earlier incarnation of this positive suggestion may be found in Winch’s talk of ‘internal relations’ at the end of the long passage quoted above. Winch further holds that reasons are akin to norms in that to act on a reason is to follow a rule of some kind. Understanding why someone acted as they did, on this view, is a matter of understanding the cultural norms which help to render the action in question intelligible. This explains both his sympathy for Collingwood’s general approach and his scepticism towards the latter’s appeal to a method of re-enactment. According to Winch, what we need to do is not share the beliefs of those whom we are trying to understand but, rather, get to grips with the socio-cultural norms which provide the normative framework within which the action took place. On the above view, understanding others doesn’t require the use of empathy as some kind of emotional instrument which provides us with an access-pass to their ‘mental contents’ but, rather, a normative immersion which enables us to find our feet with one another.38 This insight – which goes against the popular grain of understanding as a form of ‘mind-reading’39 – has been all bust lost amongst the debris of the logical connection argument. In the previous section, I suggested that this Wittgensteinian approach to the relation between reasons-explanation, rule-following, and cultural practices fits well 35 36 37 38 39 Davidson, “Actions, Reasons, and Causes”, 14 (emphasis in original). F. Stoutland, “The Logical Connection Argument”, American Philosophical Quarterly, 7 (1970), 125. Von Wright, Explanation and Understanding, 93. This argument should be distinguished from his view that actions are logically/conceptually/intrinsically connected to their results (op. cit. 87–8). Wittgenstein (1967: §383–390). See, for example, S. Nichols & S.P. Stitch, Mindreading: An Integrated Account of Pretence, Self-Awareness, and Understanding Other Minds (Oxford: Oxford University Press, 2003). journal of the philosophy of history 9 (2015) 372–392 JPH_009_03_F4-Sandis_372-392.indd 385 10/13/2015 12:31:42 PM 386 Sandis with Collingwood’s insight into historical understanding. It remains vulnerable, however, to a serious objection against the thought that reasons can explain action merely by virtue of rendering them intelligible in such ways. It is to this challenge that I now turn. III The Causal Constraint In arguing that the ‘primary’ reason for an action is its cause, Davidson was looking for a criterion for distinguishing between the reasons one actually acts upon and other reasons we might have and which we might have acted upon. Such reasons are capable of rendering our actions intelligible in the sense that they offer what he calls an interpretation of why we might have acted upon them insofar as it ‘fits it into a familiar picture’ e.g. by demonstrating that it would have been rational or appropriate to do so,40 but this form of re-movement from puzzlement is insufficient to explain why we actually acted as we did, for we may have not acted for those reasons at all: . . . a person can have a reason for an action, and perform the action, and yet this reason not be the reason why he did it. Central to the relation between a reason and an action it explains is the idea that the agent performed the action because he had the reason. Of course, we can include this idea too in justification; but then the notion of justification becomes as dark as the notion of reason until we can account for the force of that ‘because’.41 It is this point that Davidson adapts from Hempel, who states of it as follows: To show that an action was the appropriate or rational thing to have done under the circumstances is not to explain why in fact it was done . . . the 40 41 While rejecting her view that one can explain an action simply by putting it ‘in a certain light’ Davidson nonetheless follows Anscombe, Intention, p.21 in holding that showing its ‘desirability characteristic’ e.g. why it would been tempting (even if vicious or otherwise inappropriate) to perform it is sufficient for rendering an action intelligible. Hempel, by contrast, uses the phrase ‘render intelligible’ in a stricter sense which requires rationality. Nothing I say in this essay hangs on which one of them is right though it seems obvious to me that Davidson is on the side of ordinary parlance here (though he did not care much about it). Davidson (1962: 9). journal of the philosophy of history 9 (2015) 372–392 JPH_009_03_F4-Sandis_372-392.indd 386 10/13/2015 12:31:42 PM One Fell Swoop 387 presentation of an action as being appropriate to a given situation, as making sense, cannot, for purely logical reasons, serve to explain why in fact the action was taken.42 Unlike Hempel, Davidson maintains that rationalization is ‘a species of causal explanations’. By this he means that we have failed to even rationalize an action if the reason we offer for it turns out not to have been its cause. But this difference is really semantic. Davidson is using the term ‘rationalize’ as a success verb which conventionally implies successful explanation of why an event43 actually occurred. Hempel, by contrast, uses it to mean ‘render intelligible’ where this can clearly fall short of the aforementioned kind of explanation. Both philosophers agree that a reason may render an action intelligible without being its cause. This point, it seems to me, is indisputable. Consequently, there is a burden on the anti-causalist (or ‘intentionalist’ to use von Wright’s term for those who take the connection between intention and action to be logical or conceptual)44 to demonstrate what marks the difference between a reason that explains an action and a reason that (merely) renders it intelligible, if not the fact that one of these further functions as a cause. One answer to this question, to which Kenny and Melden are in different ways both reasonably sympathetic, is to defend a version of agency according to which reasons relate to actions via the agency of those who choose to act upon it. So while Melden subscribes to the ‘logical connection’ argument, it is ultimately his account of agency that enables him to state that ‘it is a fundamental mistake to suppose that the causal model employed in the natural 42 43 44 C.G. Hempel, “Reasons and Covering Laws in Historical Explanation”, as reprinted in (ed.) P. Gardiner, The Philosophy of History (Oxford: Oxford University Press, 1974), 102 & 105; see also his “The Function of General Laws in History”, Journal of Philosophy, Vol. 39. (1942), and “Rational Action”, in Proceedings and Addresses of the American Philosophical Association (1962). For now I leave aside the vexed question of whether or not actions are events. Collingwood and Winch seem happy to use the term but they arguably do so in a sense that is significantly different from Davidson’s. I argue elsewhere – in “The Explanation of Action in History”, Essays in Philosophy, Vol. 7, No. 2, June (2006) & The Things and Why We Do Them (Basingstoke: Palgrave Macmillan, 2012) Chs. 1 & 2 – that giving the reason for which one acted is an altogether different exercise from explaining why the event of their acting occurred. While this point differs in detail from those of the red book authors, I shall in due course employ such distinctions to gesture a way forward out of the impasse between Davidson and the small red books. Von Wright (1971: 95). journal of the philosophy of history 9 (2015) 372–392 JPH_009_03_F4-Sandis_372-392.indd 387 10/13/2015 12:31:42 PM 388 Sandis sciences will fit the everyday explanation of actions in terms of intentions, interest, desires, etc.’ (Melden 1961: 199). A young Annette Baier (then Stoop) took objection to Melden’s dismissal of what he termed the ‘Humean model’ of causation, rightly objecting that ‘until Melden has told us precisely what he means by a Humean cause we do not know whether he is right or wrong in maintaining that a logical connexion precludes such a causal connexion’.45 Be that as it may, Melden’s alternative positive account of a constitutivist account of agency as a non-causal form of truth-making (to which I shall briefly return below) has been unjustly neglected. A complementary line of resistance is to be found in the pluralistic account of action explanation provided by R.S. Peters who distinguishes between four different kinds of explanation: (a) ‘His’ Reason Explanations: ‘what end he had in mind (Ibid: 4) . . . his reason – whether real or not – entails that a man is conscious of his objective’ (9). (b) ‘The’ Reasons Explanations: ‘a way of calling attention to the law or assumed law that a given case falls under. His reason may coincide with the reason. The reason why Jones crossed the road might in fact be his desire for tobacco. This would then be his real reason’ (8).46 (c) Causal Explanations: ‘When there is some kind of deviation from the rule-following model . . . the appropriate answer . . . may be in terms of a causal theory’ (10).47 45 46 47 A. Stoop, “Reviews: Free Action by A.I. Melden”, Australasian Journal of Philosophy, 40:2 (1962), 248.I n her later work – e.g. “Acting in Character” in (ed.) C. Sandis, New Essays on the Explanation of Action (Basingstoke: Palgrave Macmillan, 2009) – Baier argues that Hume’s own view of action explanation fits uneasily with the Davidsonian causal model. I have argued for a similar view myself, though we do not share the very same understanding of what Hume meant by ‘cause’; see my “Hume and ‘Motivating Reasons’ ”, in (ed.) C. Pigden, Hume on Motivation and Virtue: New Essays (Basingstoke: Palgrave Macmillan, 2009), “Action, Reason, and the Passions”, in (eds.), A. Bailey and D. O’Brien, The Continuum Companion to Hume (London: Continuum, 2011), “Pouring New Wine into Old Skin: Hume’s Necessary Connexions, in (eds.) K. Allen & T. Stoneham Causation and Modern Philosophy (London: Routledge, 2011), and “A Just Medium: Empathy and Detachment in Historical Understanding”, Journal of the Philosophy of History, Volume 5, Issue 2 (2011), 179–200. An anonymous referee has suggested that it is scientistic to explicate ‘the reason explanations’ in terms of causal laws. Whether or not this is the case partly hangs on how strict the laws in question are supposed to be. But even if Melden were to embrace a form of scientism here, it is not all-encompassing for he asserts that there are other legitimate forms of explanation that cannot be reduced to this model. Cf. Melden, Free Action, 184. journal of the philosophy of history 9 (2015) 372–392 JPH_009_03_F4-Sandis_372-392.indd 388 10/13/2015 12:31:42 PM One Fell Swoop 389 (d) End-state Explanations: ‘Such explanations . . . do not give a man’s reason for eating but the reason why he eats’ (21). While we may not wish to endorse this precise way of carving things up, a pluralism along these general lines is a viable contender for best approaching the sorts of questions we’ve been pursuing. The first lesson to note is that these are not four different explanations of one and the same thing viz. action or behaviour. As Peters, maintains, ‘[t]here are many different sorts of questions which can be asked about human behaviour and the differences . . . are such that an all-embracing theory is inappropriate’ (Peters 1958:3).48 Consider, for example (a) and (b) above. Philosophers have tended to conflate these, typically subsuming both under the notion of a ‘motivating reason’ which is often taken to be interchangeable with that of an ‘explanatory reason’. In the case of Davidson, a ‘primary reason’ is meant to both motivate my action, explain it, and also justify it ‘in some anaemic sense’. Moreover it is meant to be my reason viz. a reason which I act upon and which can serve as a premise in a practical syllogism (whose conclusion Davidson takes to be an action). This is quite a lot to ask of one single entity (if that is even what reasons are) and Davidson doesn’t seem to have considered the possibility of distinguishing between ‘his reason’ explanations and ‘the reason why’ explanations of action.49 This leads to some irresolvable tensions in his account, as evidence in the following claim: Your stepping on my toes neither explains nor justifies my stepping on your toes unless I believe you stepped on my toes, but the belief alone, true or false, explains my action.50 The sentence suggests that what explains my action is not what I believe (viz. that you stepped on my toes) but my believing that you did so. Yet if we ask what would rationalize my action, the former would be a better candidate than the latter, especially if we are to follow Davidson in calling this ‘the agent’s reason’. Davidson himself fudges the issue: 48 49 50 For a very positive review of Peters’ book see R. Thomson, “The Concept of Motivation by R.S. Peters”, Philosophy Vol. 34, No. 128 (Jan 1959), 72–73. As an anonymous referee reminded me, Davidson famously leaves room for the possibility that an agent can be mistaken about what ‘his’ reasons are, but this would not be a case of ‘the’ reason differing from ‘his reason’. On the contrary, to be mistaken about what your or anybody else’s reasons are, on Davidson’s view, is to identify them with beliefs and pro-attitudes that are not ‘the’ reason for the act in question. Davidson, “Actions, Reasons, and Causes”, 8. journal of the philosophy of history 9 (2015) 372–392 JPH_009_03_F4-Sandis_372-392.indd 389 10/13/2015 12:31:42 PM 390 Sandis . . . there is a certain irreducible though somewhat anaemic sense in which every rationalization justifies: from the agent’s point of view there was, when he acted, something to be said for the action.51 The trouble is that what is to be said for the action, the sort of thing that can figure in my deliberations, is a consideration that I act upon, something that I believe to be the case, not my believing it to be so.52 Davidson switches between the following five characterisations of a ‘primary reason’ with no qualms: 1. 2. 3. 4. 5. The agent’s reason for doing what he did (p. 3). The reason that rationalizes the action. (p. 3). The reason that explains the action (p. 3). The reason why an agent did something (p. 4). The agent’s reason’s in acting (p. 11). It is far from obvious that these are interchangeable, let alone that they are all in the business of explaining one thing e.g. why A acted or why the event of her acting occurred. More importantly, for our purposes, Davidson has failed to provide philosophers like Peters with an argument for why the agent’s reasons in acting should also act as a cause of her action. Indeed, it isn’t even clear why the agent’s reason should explain her action. Davidson is looking for an all-in-one reason where there is a plurality of notions to distinguish between, and of things we might wish to explain. Each sense of reason is tied to a different sort of question we might ask about action or behaviour. Pari passu, to say that one is explaining an action, even under a certain description, is to not yet specify what it is about the action that one is wishing to explain.53 For example it is arguably one thing to ask for my reason for putting on the rubber gloves and quite another to ask why the event of my putting on the rubber gloves occurred. On a constitutivist account of action such as Melden’s, for example, one can maintain that the latter event 51 52 53 Davidson, “Actions, Reasons, and Causes”, 9. See J. Dancy, Practical Reality (Oxford: Clarendon Press, 2000), 126ff.. For the anemic sense in which the latter might be said to rationalize see M. Smith & P. Pettit, “Parfit’s ‘P’ ”, in (ed.) J. Dancy, Reading Parfit (Oxford: Blackwell, 1997), 71–95. But if this really is all Davidson has in mind he is not entitled to identify it with a reason for which the agent acts (see Sandis, The Things We Do and Why We Do Them, 73–4). This general point about explanation has been made by P. Achinstein, “The Object of Explanation”, in (ed.) S. Körner, (ed.) Explanation (Oxford: Blackwell, 1975). I relate it to action explanation in C. Sandis “The Objects of Action Explanation”, Ratio, Vol. XXV, No. 3, Sep (2012), 326–44, and the first two chapters of The Things We Do and Why We Do Them. journal of the philosophy of history 9 (2015) 372–392 JPH_009_03_F4-Sandis_372-392.indd 390 10/13/2015 12:31:42 PM One Fell Swoop 391 occurred just because I put the rubber gloves on, in the sense that my putting on the gloves is my making it the case that the event of my putting on the gloves occurred.54 This explanation of my action-event doesn’t cite my reason for putting them on. Conversely, my reason (e.g. that I was preparing to wash up the dishes) may not in itself explain why the event of my acting occurred, let alone why it did so as and when it did. Taking on board both Hempel’s and Melden’s distinctions, then, we should differentiate between at least the following four explanatory aims: 1a) Rendering what happened intelligible: explaining why an action eventor-process might have occurred. 1b) Explaining why an action event-or-process actually occurred. 2a) Rendering it intelligible why a someone acted as she did. 2b) Explaining why someone actually acted as she did. So what exactly is the difference between (1) and (2) viz. between explaining or rendering intelligible why an event of someone’s acting occurred and explaining or rendering intelligible why someone actually acted as she did? One apparent difference is that only the second set of sentences explicitly mentions an agent. But it would be begging the question against those causal theories of action which identify agents with collections of psychological states (e.g. Andrei Buckareff)55 to assume that agency is absent in the second set of examples. The difference is rather one of ontological structure: if I can make it true that my action event-or-process of raising my arm occurs by raising my arm then, upon realising this, my reason for raising my arm could just be in order to make it true that the action event-or-process of raising my arm occurs.56 Yet what explains its occurrence will not be this reason but, rather, the fact that I raise my arm. The reason why the event occurs will just be that I performed the relevant action, whereas my reason for performing the act in question will be in order to ensure that the event of my performing it occurs. 54 55 56 For a contemporary revival of Melden’s understanding of agents as truth-makers see C. Sandis, “Agents as Truth-Makers”, forthcoming. A. Buckareff, “How Does Agent-Causal Power Work?”, The Modern Schoolman, Vol. 88 (2011), 105–121. For a defense of view that the things we do or perform are not events see J. Hornsby, Actions (London: Routledge, 1980). journal of the philosophy of history 9 (2015) 372–392 JPH_009_03_F4-Sandis_372-392.indd 391 10/13/2015 12:31:42 PM 392 Sandis Epilogue The variety of nuanced distinctions relating to reasons and action to be found in the small red books point towards a pluralisitc approach which rejects the dichotomies between causalism and anti-causalism, historicism and ahistoricsm, scientisim and anti-scientism. I have tried to show that the arguments offered by Melden and Peters are not restricted to the main negative thesis of the ‘logical connection argument’ account, which Davidson rightly dismissed. The argument is at its strongest when presented more positively from within the wider framework of human understanding, as put forward by Dilthey et al. Hempel and Davidson’s objections to this move was that the resulting account of explanation was not so much false as incomplete. I have argued for an alternative means of completing it, to be found within variety of subtle conceptual distinctions that may be found within the small red books themselves. These have largely been forgotten,57 and the philosophy of action will be all the worse for it until the anti-causalist baby has been rescied from the historicist bathwater. I have defined historicism as the view that ‘historical explanation is not lawlike at all’. But if we are to espouse pluralism about action explanation this view must rejected in favour of one which allows that historical explanation can but need not be historical in character, depending on ones precise explanatory aims.58 To appreciate this is to steer clear of the Scylla of historicism, without crashing into the Charybdis of scientism. 57 58 Two exceptions are Julia Tanney (see especially her “Reasons as Non-Causal, Contextplacing Explanations”, in Sandis, New Essays on the Explanation of Action), and H.-J. Glock, “Reasons for Action: Wittgensteinian and Davidsonian Perspectives in Historical and Meta-Philosophical Context”, Nordic Wittgenstein Review 3, No. 1 (2014), 7–46. Under the influence of Anscombe, Paul Ricœur (2004:182–88), has similarly argued that there are various equally legitimate types of historical explanation, each corresponding to a different use of ‘because’ (P. Ricœur, Memory, History, Forgetting, trns. K. Blamey & D. Pellauer (Chicago: University of Chicago Press, 2004). journal of the philosophy of history 9 (2015) 372–392 JPH_009_03_F4-Sandis_372-392.indd 392 10/13/2015 12:31:42 PM