Trying to Act Together. The Power of Trust

Transcription

1 Trying to Act Together The Power of Trust

2 The Ques4on What kind of a"tude do agents have to have towards their partners in joint ac4on? How do the inten4ons of par4cipants in joint ac4on represent their partners?

3 Three Answers 1. The Cogni4vist View (agents cogni-vely expect their partners to cooperate [representa4on with mind- to- world direc4on of fit]: cogni-ve a"tude) 2. The Norma4vist View (agents norma-vely expect their partners to cooperate [representa4on with world- to- mind direc4on of fit]: norma-ve a"tude) 3. An Affec)ve Account (agents represent their partners behavior in such a way as to cogni4vely and norma4vely expect that behavior to be influenced by this representa-on: trust)

4 Raimo Tuomela s Account A member A i of a collec-ve g we- intends to do X iff (i) A i intends to do his part of X (as his part of X); (ii) A i has a belief to the effect that the joint ac4on opportuni4es for an inten4onal performance of X will obtain (or at least probably will obtain), especially that a right number of the full- fledged and adequately informed members of g, as required for the performance of X, will (or at least probably will) do their parts of X, which will under normal condi4ons result in an inten4onal joint performance of X by the par4cipants; (iii) A i believes that there is ( ) a mutual belief among the par4cipa4ng members of g ( ) to the effect that the joint ac4on opportuni4es for an inten4onal performance of X will obtain (or at least probably will obtain); (iv) (i) in part because of (ii) and (iii) (R. Tuomela, 2009)

5 Is Inten4on Belief? Focus on individual inten4on: to intend to φ is to expect (believe) that one will φ (Grice 1978). Intending with uncertainty: In order to intend to φ, A does not have to believe that he or she will φ. In wri4ng heavily on this page I may be intending to produce ten legible carbon copies. I do not know, or believe with any confidence, that I am succeeding. But if I am producing ten legible carbon copies, I am certainly doing it inten4onally. [ ] It is hard to imagine that the point does not carry over to pure intending. (Davidson 1978, 92; cf. Harman s [1986] sniper case). à Is inten4on no belief, or rather a probabilis-c belief (incl. believing with diffidence ) or some (however minimal) degree of belief?

6 Trying Beyond Belief Incompa)bility of inten)on with a conflic(ng categorical or fully certain belief: (Most philosophers think that) A cannot intend to φ if A is certain that he or she will not φ (cf., e.g., Baier 1971). But: A may s4ll try! Suppose that for some ridiculous reason Ken desires desperately to move the North leg of the St. Louis Gateway Arch one foot to the North at its founda4on, with his bare hands and no mechanical devices. [ ] Ken places his feet firmly upon the ground, gets a good grip, and pushes against the stainless steel wall of the North leg of the Arch with all of his might; all the while knowing full well that it is impossible for him to move the Arch one foot Northward. (Adams 1995, 553) Ken tries to φ without intending to φ. Imagine that the Arch actually moves. Even though Ken would be pleased with the result eventually (remember his desire), he would certainly be deeply shocked at the effect of his doing, very much unlike a person whose inten4on has just reached its condi4ons of sa4sfac4on!

7 Two Concepts of Trying Ken does not need to φ for his inten4on to reach its condi4ons of sa4sfac4on). Intending to φ and intending to try to φ have different condi)ons of sa)sfac)on - - > they are different ac)ons! But there is a sense in which any ac4on implies trying (to intend to φ, when it is inten4on- in- ac4on, is to try to φ). Confident trying: intending to φ (the intended ac4on is to φ) Diffident trying: trying to φ (the intended ac4on is to try to φ).

8 Confidence and Op4mism How can Carbon- Copy- Davidson intend to φ without any confidence that he will φ? The word confidence is ambiguous: 1) Confidence as complementary to diffidence: less- than- full confidence is a degree of diffidence; less- than- full diffidence is a degree of confidence. 2) Confidence as exclusive of diffidence (op)mism!): an agent is either confident (es4mated probability very high) or diffident (es4mated probability less than very high). à Carbon- Copy- Davidson intends to φ without any op)mism that he will φ. (If he is reasonable, Davidson must have a reason for which this seems to him to be worth trying.)

9 Condi4ons of Inten4on Confidence Condi)on: If an agent intends to φ, he or she is, to some degree, confident that he or she will φ. If he or she is fully diffident, he or she may intend to try to φ. Op)mism Condi)on: if a reasonable agent intends to φ he or she is op)mis)c that he or she will φ. If he or she is pessimis)c, but has a reason for which it is worth trying, he or she may intend to try to φ. Sufficient Reason Condi)on: a reasonable agent does not ignore reasons for confidence or diffidence / for op4mism or pessimism systema4cally (e.g., learning from experience).

10 A Cogni4vist View A reasonable member A i of a collec-ve g we- intends to do X iff: (i) A i intends to do his or her part of X (as his part of X); (ii) A i is op)mis)c that g will X, i.e. that he or she and the relevant others will do their parts of X. (iii) A i is op)mis)c and has no reason for pessimism that g will X; he or she does not have to take his or her op4mism to be universally shared, let alone to be common knowledge, as long as he or she assumes that the pessimis)c and the totally diffident members of g will try (diffidently) to do their parts of X successfully. (iv) (i) in part because of (ii) and (iii) (No mutual belief required; joint ac4on may include totally diffident and pessimis4c [as well as indifferent] group members.)

11 Puong Cogni4vism to the Test

12 The Test Case November 14, 1951 Charles M. Schulz Football Gag November 16, 1952

13 50 Years of Geong Suckered : Charlie Brown killing himself (his own words)

14 What Does Charlie Think he s Doing?

15 The Cogni4vist Verdict Charlie s reac4on at the end of most rounds clearly shows that his trying is of the op)mis)c kind: his inten4on is to kick the ball rather than just to try (diffidently) to kick the ball (his inten4on clearly does not reach its condi4ons of sa4sfac4on: he is surprised and disappointed). Charlie s inten4on to kick the ball represents Lucy (cogni4vely) as doing her part (holding the ball steady). Charlie and Lucy do not grow older, but this does not mean that Charlie does not accumulate experience (he osen refers to past rounds explicitly). Already in the 1950ies, it becomes increasingly clear that Lucy will not hold the ball steady. This evidence is further corroborated through the 60ies, 70ies, and 80ies. Given the evidence, Charlie has ample reason not be op)mis)c. By con4nuing to intend to kick the ball, Charlie violates the sufficient reason condi)on.

16 What the Public Thinks The fault isn t Charlie s. Lucy rather than Charlie has reason to change her attude! This may not be the most reasonable advice; but it seems wrong to say that it is always unreasonable to trust in the face of past contrary experience.

17 A Norma4vist Account Individuals A 1..A n form a plural subject of X- ing (for some ac4on X or psychological atribute X) if and only if A 1...A n form a joint commitment to X- ing as a body. Joint commitments are formed when each of a number of people (implicitly) expresses his or her willingness to par4cipate in the relevant joint commitment with the others. Once the joint commitment is established, each individual is individually obligated to do his or her part to make it the case that he or she acts according to the joint commitment, and has a right to expect the same from the other par4cipants. (M. Gilbert) The par)cipants expect their partner s behavior norma)vely rather than cogni)vely

18 Norma-ve Expecta4ons Inverted Direc)on of fit/direc)on of responsibility : mind- to world (cogni4ve expecta4on); world- to- mind (norma4ve expecta4on) (inverse direc4on of blame in the case of disappointment) Limited range of possible objects: to expect norma4vely is to conceive of the addressee as suscep)ble to norma)ve reasons, i.e. to be, to some degree, confident that under suitable circumstances, he or she will act accordingly (ought implies can). Ra)onal immunity to disconfirma)on ( counterfactual stability [Luhmann]; just that A did not φ, and that it is likely that he will not φ, does not mean that it was wrong to norma)vely expect A to φ) à past disappointments do not per se make an agent s confidence in another agent unjus4fied.

19 Defending Charlie Insofar as Charlie is jus4fied in assuming that Lucy is a person (a creature who is suscep4ble to norma4ve reasons) and knows what she ought to do, and that she ought to hold the ball steady, he has sufficient reason for being confident about her holding the ball steady, and thus for trying to kick the ball. This confidence is such that it is not ra4onally undermined by past contrary experience. à Charlie is jus)fied in not giving up!

20 Puong Norma4vism to the Test

21 Norma4vism Fails, too! 1. The claim Charlie has sufficient reason to intend to do his part because of his agreement with Lucy begs the ques4on of whether there is sufficient reason for Charlie to enter into an agreement with Lucy. (The norma4vist response to this ques4on sets off an infinite regress [cf. Charles M. Schulz meta- agreements].) 2. Norma4ve expecta4ons ra4onalize diffident trying, but not confident trying: While it seems true that Charlie has sufficient reason for confidence (ought implies can), he clearly does not have reasons for op)mism. It seems that given the evidence, his norma4ve expecta4on licenses him only to intend to try to do his part (diffidently). à Insofar as he con4nues to intend to kick the ball rather than just intending to try to kick the ball, he has insufficient reason and is unreasonable.

22 Taking Stock Neither the cogni)vist nor the norma)vist view can accommodate the intui)on that Charlie s con)nued confident trying to hit the ball is not (en)rely) unreasonable. Is the intui4on wrong, or is there something wrong with the views? An indica4on towards the later: The (pure) cogni4vist view misconstrues of coopera4ve- mindedness because it is blind for the role of norma)ve expecta)ons in coopera4on; the (pure) norma4vist view misconstrues of coopera4ve- mindedness because it is blind for the reasonableness constraints placed by cogni)ve expecta)ons. à How can the two views be combined? How do cogni)ve and norma)ve expecta)ons work together?

23 Three Steps Towards an Alterna4ve Account 1. At the basic level, cooperators represent their partners as doing their parts in the same way as individual inten4ons implicitly represent the agent as con4nuing to be willing and able to perform the ac4on un4l the inten4on s condi4ons of sa4sfac4on are reached (individual agents of temporally extended ac4ons represent their own future behavior in the same way in which cooperators represent their partners behavior). 2. An individual agent s representa4on of his own future behavior (and self) is neither purely cogni4ve nor purely norma4ve (an individual with a purely cogni4ve stance towards his own future self s behavior and no norma4ve expecta4on is a predictor of his behavior rather than an intender of his future ac4on; similarly, an individual with a purely norma4ve stance towards his own future behavior is a valuator his future behavior rather than an agent). 3. In individual inten4on, cogni4on (predic4on) and norma4ve expecta4on are related in such a way that the agent takes himself to be jus)fied to assume that he will con4nue the intended course of ac4on in the future partly because the current inten4on that represents him as doing so provides him, in the future, with a (further) norma)ve and mo)va)ng reason to do so and therefore makes it more likely that he will do so (bootstrapping). The kind of representa4on that does this is self- trust.

24 The Power of Trust In coopera4on, interpersonal trust works in the same way as self- trust (self- confidence) does in the case of individual ac4on: A cooperator s basic aotude towards his partner is such that he (implicitly) assumes that by represen4ng the other as doing his part he makes it more likely that the other will in fact do his part because it provides the other with a mo4va4ng and a norma4ve reason to do so (the Power of Trust). The Power of Trust is based on the addressee s trust- responsiveness (PeTt) (à psychol. cost of betraying trust) as well as his assumed trust- suscep)veness (being actually relied upon adds a further norma)ve reason to the fact that one is under an obliga4on). The Audacity of Trust: Because of the partly self- fulfilling nature of trust, the limits for ra4onal trust are difficult (or perhaps impossible) to ascertain (on the background of nega4ve past experience, a bold display of trust may well be assumed to be par-cularly effec-ve; cf. the public s nega4ve response to L.).

25 Trying to Act Together What Charlie intends to do is to try to act together (with Lucy) rather than to act together (with Lucy). Intending to try to act together with Lucy is to do whatever is necessary to act together with Lucy without being op4mis4c about the outcome. Since ac4ng together involves trust, Charlie has to trust Lucy in order to be able to intend to try to act together. Therefore, a condi)on of his intending to try to act together with Lucy is his op)mis)c trying to kick the ball (he cannot just pessimis)cally intend to try to kick the ball, since this does not represent Lucy as doing her part and thus does not have the power of trust). In order to be a proper intender of the trying to act jointly, he must appear, under the given circumstances, to be a unreasonable par)cipant in the coopera4on.

26 Did the Power of Trust Ul4mately Prevail? October 24, 1999: Charlie s Last Run