Abstract: Do we really care whether our beliefs are true? Stephen Stich gives us a very surprising but challenging answer: once we find out what it means for a belief to be true, the answer to the above question is “a consistently negative” one. He argues that there is neither intrinsic nor instrumental value in having true beliefs. However, his argument is based on some very dubious reasons. For instance, one of his reasons is that if we value true beliefs intrinsically, we will leave out a huge space of mental states that have no truth values but would vastly increase their user’s power or happiness or biological fitness. But this is false because we can value different things intrinsically at the same time. He is even less successful in arguing against instrumental value in having true beliefs. He admits that he does not establish a knockdown argument against the value of having true beliefs, but he insists that the burden of argument be surely on those who maintain that there is value in having true beliefs. To meet his challenge, we have shown that there is cognitive intrinsic value in holding true beliefs and that generally, true beliefs are more conducive to our survival than false beliefs. If we completely depend on our false beliefs to achieve our goals, we will act like a blind cat who can only catch a mouse by chance.
Keywords: truth, true belief, intrinsic value, instrumental value, pragmatism
Do we really care if what we believe is true? This seems to be a trivial question to many of us, if not all of us, since the answer is obvious: we do care about having true beliefs. So many examples in our ordinary lives can be used to support this answer. If I don’t want to miss a date with my girlfriend, I do care whether I remember the time of the date correctly. If I don’t want my daughter to be hurt by a hot stove, I do care whether she forms a relevant true belief about it. However, in his book, The Fragmentation of Reason, Stephen Stich gives us a very surprising but challenging answer: once we find out what it means for a belief to be true, the answer to the above question is “a consistently negative” one. (Stich 1990, p. 22)
In this paper, I will defend the value of having true beliefs. In my view, it is sufficient to show that we do care if what we believe is true if any one of the following theses is established:
I. There is intrinsic value in having true beliefs.
II. There is instrumental value in having true beliefs.
Thus, to show that we don’t really care about truth, it is necessary to establish both of the following:
NI. There is no intrinsic value in having true beliefs.
NII. There is no instrumental value in having true beliefs.
Actually, these two theses are the ones Stephen Stich tries to establish in support of his position. He says, “once we have a clear view of the matter, most of us will not find any value, either intrinsic or instrumental, in having true beliefs.” (Ibid., p. 101) In this paper I will show that he does not establish NI and NII. Stich admits that he does not establish a knockdown argument against the value of having true beliefs, but he insists that the burden of argument be surely on those who maintain that there is value in having true beliefs. (Ibid., p. 124) To meet his challenge, I will deny theses NI and NII by establishing theses I and II.
I
In this section, I will discuss whether Stich has succeeded in establishing thesis NI. According to some subjective notion of intrinsic value, something is intrinsically valuable if someone desires it intrinsically (see a more detailed discussion in the second section of this paper). If this were the case, it would be almost impossible for Stich to establish NI because it is easy to find someone who values true beliefs intrinsically. Stich seems to be aware of such a difficulty, but he insists that nobody should attach intrinsic value to truth. What he actually defends is the following thesis:
NI1. We should not attach intrinsic value to holding true beliefs.
Stich argues that once we are clear about what it is for a belief to be true, we will find out that valuing truth intrinsically is “a very conservative thing to do,” i.e., those who place value on having true beliefs are placing an unjustifiable importance on some particular way of mapping mental states onto the extra-mental world. [2]
What is it for a belief to be true? According to Stich, a relatively plausible theory of truth is what he calls “the causal/functional theory.” (Stich 1990, p. 106) This theory is actually a combination of Taski’s theory of truth and the causal theory of reference for names and kind terms. Based upon this theory, a belief as a brain state or a complex neural event does not have semantic properties (i.e., truth values). In order for it to have a truth value, it has to be mapped onto a proposition, P. Thus a belief is true if and only if P is true. P is true if and only if its truth conditions or some possible states of affairs obtain. Some possible states of affairs obtain if and only if there is a coherent, causal and commonsense story about the mapping relation between the belief and the possible states of affairs. Therefore, according to the causal/functional theory, a belief is true if and only if it is paired up with its truth conditions through a coherent, causal and commonsense interpretation. Stich calls this interpretation mapping “the causal/functional interpretation function.” (Ibid., pp. 22-23, 103-104, 107-110)
Now we can formulate Stich’s argument for NI1 as follows:
(1) If the only plausible way of determining truth is the way specified by the causal/functional theory, then we should not attach intrinsic value to having true beliefs.
(2) The only plausible way of determining truth is the way specified by the causal/functional theory.
(3) Therefore, we should not attach intrinsic value to having true beliefs.
Both premise (1) and premise (2) seem to be questionable. Let’s consider premise (1) first. One of Stich’s reasons for premise (1) is that the causal/functional theory or “the causal/functional interpretation function” leaves out “many possible [useful] systems of mental states to which no interpretation will be assigned, either because the elements out of which those states are constructed are not hooked on to the world by the appropriate sort of causal chain or because the formal patterns of interaction [which] the compounds manifest admit of no intuitive semantic interpretation [For example, there is no intuitive semantic interpretation of the quantifiers and connectives of a language.].” (Ibid., pp. 113-114) Thus, those who would accord intrinsic value to the holding of true beliefs may well be reluctant to explore a huge space of many systems of mental states that have no truth values but would vastly increase their user’s power or happiness or biological fitness. (Ibid., p. 119) Therefore, premise (1) is true. However, this reason is false. For the person who values having true beliefs intrinsically may also value other things intrinsically, and thus valuing truth intrinsically does not necessarily prevent him from exploring mental states that have no truth values but may contribute to some other intrinsically valuable ends.
Another reason Stich uses to support premise (1) is that that if we map our mental states to truth conditions only through commonsense intuition, we simply let our traditions (commonsense intuitions) to determine our cognitive values without critical evaluation of those traditions. (Ibid., p. 120) Even in the domain where it can specify interpretations, there are many other functions that map mental states to TRUTH CONDITIONS such as TRUTH* CONDITIONS, TRUTH** CONDITIONS, etc., which are not sanctioned by commonsense intuition. And “there is nothing obviously superior or preferable about the one sanctioned by commonsense intuition.” (Ibid., p. 114) But Stich’s reason does not work. First, his reason assumes that we formed our tradition without critical thinking, but this assumption seems to be false because we (especially philosophers) critically examine our traditions all the time. Second, his reason assumes that our traditions cannot be changed, but they can be changed due to new evidence, new discovery, new technology, etc. as well as critical thinking. Third, we value many things other than truth intrinsically, such as survival, pleasure, peace, beauty, etc. Thus, we actually allow some other intrinsically valuable things to help “correct” our traditions. Therefore, Stich does not establish the truth of the first premise.
Now I want to argue that the second premise is also false. Stich’s criticism of the causal/functional theory of truth, at best, is about the problems of that theory, but not about how we actually determine whether a belief is true. To determine whether a belief is true is one thing, and the theory of how we determine it is true or what truth means is another. We knew whether our beliefs are true long before we had any theory of truth. Thus, even if a theory of truth has some problem, this does not mean that we cannot determine whether our beliefs are true. For instance, we can determine whether “killing an innocent person for no good reason is morally wrong” or “2 + 2 = 4” is true without knowing any theory of truth. We don’t need to understand some theory of truth in order to determine what is true and what is false, as long as “no one could reasonably reject.” (Scanlon 1998, p. 153) Therefore, Stich’s criticism is, at most, a criticism of a certain theory of truth but not of how we determine the truth.
Since we have many ways to determine or believe something is true, not limited to the ways defined by the causal/functional theory, even if Stich were correct, he only showed the problems of the causal/functional theory not the problems of our ways of determining whether our beliefs are true. Thus, the second premise is false. Therefore, the conclusion does not follow.
II
Even though, as I have shown above, Stich does not establish NI or NI1, this does not necessarily mean that there is intrinsic value in having true beliefs. We need some positive reasons for the intrinsic value in holding true beliefs. In this section, I want to defend the intrinsic value, especially the cognitive intrinsic value, in holding true beliefs, based upon three possible definitions of intrinsic value.
There are at least three possible definitions of intrinsic value: subjective, objective and counterfactual. According to the subjective definition, X is intrinsically valuable if and only if someone values it for its own sake, [3] regardless of the consequences of valuing it. If this is what intrinsic value means, then it seems to me that it is difficult to deny that true beliefs have intrinsic value, since to show this is true is simply to find someone who intrinsically values having true beliefs. Many people value true beliefs for their own sake. For instance, Giordano Bruno (1548-1600) seemed to value true beliefs for their own sake. Bruno did not give up his belief in the truth of Copernicus’s cosmology even though holding such a belief led to his being burned at the stake by the Inquisition. David Hume believed that people’s intrinsic desires could not be rationally justified or rejected. Reason is incapable of proving or disproving rationality of intrinsic desires. If this were the case, and if intrinsic value is determined by intrinsic desires, it would be almost impossible for Stich to deny that having true beliefs is intrinsically valuable.
However, Stich may disagree with Hume and the subjective definition of intrinsic value. He may argue that there is no intrinsic value in having true beliefs, no matter whether someone thinks there is or not. This seems to assume an objective external definition of intrinsic value, according to which, X has intrinsic value if and only if it is a fact that X has value and its value in no way depends on anything (anyone or any desire) else in the world. [4] It is very controversial whether there is such a value. The problem is: how do we know or determine whether X has such intrinsic value or not? Or according to what standards can we determine whether X has such intrinsic value? If we could not find a way or a test to determine whether X has such value or whether such value exists, then we would be not only unable to avoid some ontological and epistemological worries about the queerness of such values, as raised by John L. Mackie (See Mackie 1977, chapter 1), but also unable to prove or disprove whether there is such intrinsic value in having true beliefs. If this were the case, the debate on intrinsic value of true beliefs would lead us nowhere. [5]
Some philosophers believe that there are objective values, including intrinsic values as defined above, and if we are fully rational, once we discover such a value, we will necessarily desire it, and if such a value is intrinsic, we will desire it as an intrinsic end. (Cf. Bond 1983, chapters 4 and 5) This leads to the objective internal definition of intrinsic value or the counterfactual definition of intrinsic value, according to which, X has intrinsic value if and only if a fully rational person would desire it for its own sake, regardless of the consequences of desiring it, when he or she were well informed and ideally situated. (Cf. Railton 1986a, pp. 173-177 and 1986b, p. 9) This seems to indicate some way to determine an objective intrinsic value—we can appeal to what a fully rational person would desire or not desire in the ideal situation to determine whether something is in fact intrinsically valuable or not. It is not difficult to imagine that, even if Bruno were ideally situated and well informed, he would still hold true beliefs to death rather than give them up for survival. It is difficult for us to blame him for not giving up what he believed true even though his beliefs led him to death. Consider another case. At the trial in 399 B.C., Socrates tried to defend his philosophical activities of looking for the truth (finding out whether the oracle at Delphi was true, which said that no one was wiser than Socrates). Obviously, he was looking for the truth, regardless of the consequences and even if he was well informed and ideally situated as he was at the trial, he by no means regretted what he was doing. We cannot blame him for irrationality. This seems to be another example of showing truth has intrinsic value. Actually, we seem to believe that looking for truth is a morally praise-worthy virtue. Someone might say this seems to prove that looking for truth is not intrinsically valuable but instrumentally valuable, i.e., valuable as a means to making people better. But why can looking for truth make people morally better? Why do we praise those people who are looking for truth? In many cases, we praise them because they value truth intrinsically, regardless of the consequences of valuing it. In other words, we believe looking for and holding truth is a virtue because it is valuable for its own sake and it is worthy of being pursued. Therefore, according to the third definition, it is still difficult to deny that there is intrinsic value in having true beliefs.
Still, Stich may not be convinced by the argument above. He may say that in those examples, an intrinsic value still depends on the desire of a person, even though a fully rational person in the ideal situation. We need something other than desires to justify the intrinsic value in holding true beliefs. After all, as David Brink says, “It is not that these things [desired objects] are valuable because we desire them; rather, we desire them because we think them valuable.” (Brink 1989, p. 225) Rational people desire valuable things for reasons. They value true beliefs because true beliefs are worth believing. They value true beliefs intrinsically because they cannot give up believing what they believe true. In other words, there is the cognitive intrinsic value in holding true beliefs. In the rest of this section, I will defend the cognitive intrinsic value of having true beliefs.
One interpretation of intrinsic value could be that it couldn’t be given up even when facing terrible consequences. A belief may practically be useless to me and thus it is less conducive to my survival than many competing beliefs but cognitively I still care whether it is true. This is because cognitively people value having true beliefs intrinsically, regardless of whether it is good or not in action. For instance, my belief that there were dinosaurs on the Earth in very ancient times may be useless for me but cognitively I still care about whether it is true.
Someone may say, “What if it turns out that anyone who believes that it is true that there were dinosaurs on the Earth in very ancient times dies of a brain hemorrhage one year later?”[6] Whether this could be a true situation is in question. Let us assume it is. Now suppose I have observed fossils of dinosaurs and have relevant geological and biological knowledge, which have made me believe that dinosaurs existed. If one told me that if I still believe that dinosaurs existed I would die one year later; I believe this is true and it is true, can I truly give up my belief? If someone says to me, “If you still believe that Barack Obama is the President of the United States, you will be killed.” Suppose what he says is true and I believe it is true. I want to survive, and thus I truly want to give up that belief. I can say, “I give up the belief.” But from inside of my heart, I feel it is difficult and almost impossible not to believe the truth unless someone gives me a medicine to cause amnesia, or removes a part of my brain by surgery, or convinces me that all the information that caused me to have that belief is completely false. The Jesuit cardinal Robert Bellarmine instructed Galileo (1564-1642) that he must no longer hold the concept that the Earth moves. Otherwise he would be punished. Galileo knew how Giordano Bruno was burned at the stake and he did not want to die. He chose to remain silent for years. However, I doubted that he truly gave up his belief that the Earth moves. Why is it so difficult or almost impossible for me (or someone else) to give up some belief I have good reasons to believe true? Why is it so difficult, if not impossible, for me to give up a true belief even though it might be true that holding that true belief is not good for me? The reason is, or this fact shows, that the cognitive value of holding true beliefs is intrinsic, regardless of the consequences of holding them. Therefore, even if practically speaking a true belief is not useful for us, cognitively we still care about it.
Stich’s problem is that he seems to try to deny the cognitive intrinsic value of having true beliefs (that is why he insists that we don’t care whether our beliefs are true) or replace it with practical value in action (that is why he emphasizes that what we really care is only what is useful). But as I have shown above, this does not work. The former cannot be reduced to the latter. He made the same mistake as Pascal’s Wager did—he confuses what is true with what is useful.
III
Now let us consider whether having true beliefs is instrumentally valuable. It seems to be trivial to prove thesis II since the realization of so many ends in our ordinary lives, such as not being hurt by a snake, not missing a date with my girlfriend, etc., depends on whether our relevant beliefs are true. Thus the task of establishing thesis NII seems very difficult. As Stich himself puts it, “For to demonstrate this would require showing that true beliefs don’t facilitate anything that people might sensibly value.” (Stich 1990, p. 121) Stich admits that no one seriously attempts to prove thesis NII and he does not want to do either. However, he argues that the instrumental value of having true beliefs is far from obvious. He uses Harry’s example to illustrate his point. Harry had a true belief that his flight left at 7:45 a.m. Based upon on this belief, he got to the airport on time. Unfortunately, the plane crashed later and Harry died. If Harry falsely believed that the plane departed at 8:45, he would have missed the plane and survived. (Ibid., p. 123) Stich’s argument through this example seems to be that what we care about is not whether our beliefs are true, but rather, whether our beliefs are useful. If true beliefs had instrumental value by themselves, we would always care about them. But as Harry’s case has shown, this is not the case. Therefore, there is no instrumental value in true beliefs or having them.
But this argument can be used against instrumental value in any belief, no matter whether it is true or false. If this were Stich’s point, then, from an instrumental standpoint alone, we would not care about any belief or any mental state at all. However, arguing in this way is misleading. The instrumental value of a belief always depends on its contribution to a certain goal of action. A belief itself does not have any instrumental value unless it is related to some end the agent or believer desires. Harry’s holding a true belief or a false belief about the departure time is not useful or harmful unless he has a desire that can motivate him to act on that belief. If we take all the ends off the table, of course, there will be no instrumental value in any belief or holding any belief. We cannot talk about instrumental value of beliefs without talking about relevant ends. Stich’s actual point or the real issue is whether true beliefs are always better to direct us to achieve our goals than false beliefs.
I want to argue that true beliefs are better to direct us to achieve our goals than false beliefs. If we care about our ends, we do care about the means. If we care about the means, we do care about our beliefs about the means or our beliefs about how to achieve our ends, because, as human beings, we cannot act like a blind cat who can only catch a dead or alive mouse by chance, and our beliefs about the means can direct us to achieve our goals. It seems obvious that true beliefs about the means are better to direct us to achieve our goals than false beliefs. Therefore, from an instrumental standpoint alone, we should care about whether our beliefs are true. It seems to me this is a sound argument. Stich’s example actually challenges the truth of the statement that it is obvious that true beliefs about the means better serve our goals than false beliefs.
Consider Harry’s case again. Obviously, Harry’s true belief about the departure time will better guide him in catching the flight on time than his false beliefs. Someone may argue that this is not always the case. For example, suppose that Harry had to leave for the airport at 6:30 a.m. in order to catch the flight on time. Suppose a traffic accident happened on the way to the airport at about 6:35 a.m., and it caused a traffic jam for an hour. If Harry had a false belief that the flight left at 6:45 a.m., then he could avoid that traffic jam and catch the flight. In this case that false belief better served his goal than his true belief. But from this, it does not follow that Harry should not care about his true belief about the departure time or he should be blamed for having that true belief. First, what caused him to fail to achieve his goal is not his true belief about the departure time but, rather, an unexpected traffic accident. Second, his true belief failed to help him achieve the goal is not because that belief itself, but because he lacked another true belief about some unexpected event—unexpected traffic accident, i.e., he lacked true beliefs about the whole picture of the event. Therefore, what we can draw from this scenario (or example) is not that we should not care about true beliefs, but that we should care more about our true beliefs about the whole picture of the event, rather than about a part of the picture. Finally, the realization of the instrumental value of a false belief is by accident but true beliefs, especially true beliefs about the whole picture of the life can help us achieve our goals for sure. If we completely depend on our false beliefs to achieve our goals, we will act like a blind cat who can only catch a mouse by chance.
Similarly, relative to Harry’s goal of surviving, his true beliefs about the means to his survival will better serve his purpose than false beliefs. Since his goal of surviving is more general than his goal of catching the plane, and it might conflict with some of his other goals, more true beliefs about how to achieve that goal are required. In Harry’s case, those true beliefs should include at least true beliefs about the possible conflicts between his desire to survive and his other desires such as the desire to catch the flight. Also, what caused his death is not his true belief about the departure time but the unexpected plane crash or his false belief about the possibility of the crash or his lack of true beliefs about the whole picture of the story. As human beings, we live and act on purpose. If we don’t want to live and act like a blind cat, we need beliefs to direct our actions. However, false beliefs cannot perform such a function. Harry’s false belief about the departure time did not guide him to avoid the crash but only led him to survive by accident. On the other hand, his true beliefs about the whole picture would direct him to avoid the crash for sure, if it is possible for him to get the whole picture.
Stich may argue that, since it is impossible for Harry to get the whole picture, his true beliefs about the part of the story he knows would guide him to act in a way that would lead to his death for sure, but his false belief, at least, would save his life, though by accident. Therefore, Harry should not care about his true belief about the departure time. But first, even though it is true that his false belief could save his life, but his true belief couldn’t, it does not follow that Harry should not care about his true belief about the departure time or that he should be blamed for acting on that belief. If the principle of “ought” implies “can” (i.e., we cannot blame someone for not doing things that he is incapable of doing) is correct and if Harry didn’t know and couldn’t know the plane would crash, why should he be blamed for acting on his true belief? In other words, why should we think that it is irrational for Harry to act on his true belief? The point is that we have to act on the basis of the information available to us. We cannot blame someone for doing things that he is incapable of avoiding.
Second, if we do not care about true beliefs, i.e., if we do not follow what we believe is true (though it may be false), the instrumental value of a false belief cannot be realized. It is not difficult to imagine a rational person can act on a false belief that he believes is true, but it is difficult to imagine that he can act on a false belief that he believes is false. We can imagine that Harry could miss the flight by acting on his false belief that he believed was true but it is difficult for us to imagine that he could miss the flight by acting on his false belief that he believed was false. Therefore, the realization of the instrumental value of a false belief depends on whether we do care about truth, and it can be realized only by accident or miracle.
We should also note that “We don’t care about one particular true belief” does not entail that “We don’t care about true beliefs (true beliefs in general).” That one true belief is less conducive to our survival does not mean that true beliefs in general are less conducive to our survival. Therefore, even if Stich’s case is sound, he does not establish that we don’t care whether our beliefs are true in general. Stich is aware of this but he insists, “to show this requires an argument, and as far as I know, no one has any inkling of how that argument might go.” (Ibid., p. 124)
Now here is my argument. How can having true beliefs be better in general than having false beliefs for achieving goals? It is true that false beliefs are more conducive to our goals in some cases. But the instrumental value of having false beliefs in those cases depends on some unknown event, e, that has not happened yet but might happen or might not happen. And usually the probability of the occurrence of e is lower than that of the occurrence of ~e. Thus the instrumental value of having false beliefs is uncertain. In Harry’s case, the instrumental value of his false belief about the departure time in respect of his survival depends on some unknown event, i.e., the plane crash. If the plane crash never happened, the instrumental value of his false belief couldn’t be realized. On the other hand, the instrumental value of having true beliefs is more certain than that of having false beliefs. Failure to realize the instrumental value of having true beliefs is usually caused by some accidental events whose probabilities of occurrence are very low. We can imagine many cases in which false beliefs are more conducive to our survival. It is possible that I could run into a fatal traffic accident on the way to a restaurant where I had a date with my girlfriend, if I had a true belief about the time of the date. Such a fatal accident could have been avoided, if I had remembered incorrectly the time of the date. It is possible that I could be hit by an unexpected meteorite from the sky on the way to the library, if I remembered correctly the route to the library, but such an accident could be avoided, if I remembered incorrectly. But the probabilities of those events are very low. If we take those accidental events too seriously, we cannot do anything, because even the most innocuous activity like sitting at home could be fraught with unexpected danger: A meteorite could fall from the sky and crash through the roof upon us.
Generally speaking, the probabilities of those events occurring upon which the instrumental value of false beliefs depends are very low; the instrumental value of having false beliefs is not as certain as that of having true beliefs. Therefore, the instrumental value of having true beliefs is better than that of having false beliefs and true beliefs do a better job of helping us achieve our goals than false beliefs. In other words, false beliefs cannot guarantee the attainment of our goals as firmly as our true beliefs can. But we want to achieve our goals with certainty. Therefore, we do care about truth and about whether or not what we believe is true.
Notes
[1] I am very much indebted to the comments on the previous draft of this paper, made by an anonymous reviewer in Michigan, the United States of America. I am also grateful to Laura Tang for reading and correcting the previous version of this paper.
[2] I owe this interpretation of Stich’s phrase to the anonymous reviewer in Michigan.
[3] “S values X for its own sake” is logically equivalent to those statements: “S values X not as a means to something else” or “S values X no matter what has happened and what will happen in the world.”
[4] I owe the idea of this definition to the anonymous reviewer in Michigan.
[5] Stich seems to recognize such a problem with the objective external definition. He seems not to reject a justified subjective sense of intrinsic value and he might accept the counterfactual sense of intrinsic value. (Stich 1990, p. 93, 118)
[6] I owe this question to the anonymous reviewer in Michigan.
References
Bond, E.J. (1983). Reason and Value, Cambridge: Cambridge University Press
Brink, D. (1989). Moral Realism and the Foundations of Ethics, Cambridge: Cambridge University Press
Mackie, J.L. (1977). Inventing Right and Wrong, London: Penguin Books
Railton, P. (1986a). “Moral Realism,” The Philosophical Review, XCV, No. 2
Railton, P. (1986b). “Facts and Values,” Philosophical Topics, Vol.14
Scanlon, T.M. (1998). What We Owe to Each Other, Cambridge, Mass.: Harvard University Press
Stich, S. (1990). The Fragmentation of Reason, Cambridge, Mass.: The MIT Press
CHEN Zhen (*)
Department of Philosophy and Research Institute for Moral Education, Nanjing Normal University，Nanjing 210046, China
E-mail: chenzhen3488@sina.com
（Forthcoming in Frontiers of Philosophy in China, Number 1, 2012。录入编辑：红珊瑚）
