2003-04-05 - 4:07 a.m.

���F�D�w���ǡA�ڹ�F�v���ǡB�޿���޿���ǡB���ѽפΦs�b�D�q���]�ܦ�����A���L�b�o�Ǥ譱���o�H���峹���h�C�H�U�O�@�g�g�o���n���A���e�����`�A�A�X��Ǧ��S�����쪺�H�ݡC

Memory belief, justification and knowledge

In his article ��Memorial Justification��(1) and his later book Epistemology: A contemporary introduction to the theory of knowledge , (2)Robert Audi claims that there may be cases that we can have a piece of knowledge which is not justified to be believed. Such claim rules out the traditional idea that justification is a necessary condition for knowledge. This essay aims at examining the cogency of his argument and I will argue that Audi��s argument, though impressive and convincing, is a mistaken one. I will first start by clarifying the key concepts in his argument, and then go on to point out that his argument bases utterly on an ambiguous application of ��remembering��.

The notion of ��memory belief�� is central in Audi��s argument and it seems necessary to have a clearer concept of it for understanding his argument. By ��memory belief��, Audi means a belief grounded on memory.(3) This does not commit him to say anything about what the content of the belief is and how the belief is caused. After all, the belief grounded on memory can be a belief about past events, general truths like mathematical and logical truths and even future events. Also, while every memory belief is at least partly caused by a past event, it seems that it is not the case that every belief at least partly caused by a past event is a memory belief, as shown by his poison example; and neither every memory belief is caused by the past event stored and contained by that memory belief. These examples show that one necessary condition for a memory belief is that it must be grounded on memory�Xthe mental storehouse and capacity for beliefs and knowledge. Even if a belief is caused by a past event or if it is a belief about past events, it is not necessary a memory belief if it is not grounded on memory. Audi carefully distinguishes grounding of a belief with the cause and content of a belief in footnote number 2 in his chapter ��Memory�� in Epistemology(4) , and assigns grounding a crucial role in determining whether a belief is justified. Therefore, a belief is a memory belief only if it derives its justification from our mental capacity of memory. Here, it brings out an important distinction between (a) a memory belief and (b) a belief retained in memory. (5) A belief retained in memory is a belief we store in the mental storehouse of memory, and is ready for us to recall or remember. It is not necessary a memory belief because such belief can be grounded on whatever justification else. For example, I have a belief that I am writing a philosophy essay on the ground that I perceive that I am doing so. Then it can be retained in my memory, ready for my recalling next time. Such belief is not a memory belief because it is not grounded and justified by our memory. The retained belief may become a memory belief when I activate my memory, recalling or remembering it from the mental storehouse, and derives its justification from my recalling and remembering to make it a piece of knowledge for me. But one important point to be noted here is that my original retained belief does not derive its justification to be a knowledge from my memory, but rather on something else, namely perception in this example. It means that if there is a certain piece of knowledge retained in our memory, it does not necessarily derive its justification from memory. Only if it is remembered or recalled from memory and becomes a memory belief that it will be grounded on memory. Later on, we will see how the confusion between these two kinds of belief makes Audi to accept his mistaken conclusion.

So much with memory belief and it is time to turn to another important notion�Xremembering. For Audi, remembering is the mental capacity of having retained beliefs in the mental storehouse available for recall, as contrast with recalling as the capacity of bringing actively these beliefs back to mind. (6) Just like seeing, hearing, and feeling in the case of perception that produce perceptual beliefs, remembering produce memory beliefs by activating the beliefs retained in our memory. One important feature of remembering is that we can never remember something that is false. We cannot remember that the Second World War broke out in 1914 when it in fact broke out in 1939. The target of remembering, that is the retained belief to be remembered, must be something true. This brings out an important distinction between remembering and memory beliefs. As I have mentioned in the above paragraph, there is no restriction on the content of a memory belief, provided that it is grounded on memory even if it is false. It could happen because we can remember falsely even if what is remembered must be true. For example, if I remember that my girlfriend��s birthday is on 10th December, then what is remembered, namely her birthday is on 10th December, must be true. But it is possible for me to remember falsely and thereby produce a memory belief that her birthday is on 1st December, which is false. The importance of this distinction is that what is remembered must be true but a memory belief derived from remembering may not be true. We cannot remember something that is false, but we can remember falsely.

On the other hand, a retained belief to be remembered must also be justified too, and we cannot remember something that is not justified to believe. I cannot remember that I have a serious fever when I was 2, when I can only recall some fabric and unconnected images about my being sent to hospital, even when it is in fact true. As said by Audi, ��only properly grounded true beliefs stored there (in memory) constitute remembering�� (7)

We can now examine how Audi constructs his argument that we can have a piece of knowledge without such piece of knowledge being justified. Two sets of premises are used here, namely that: (1) remembering something entails knowing something, and (2) remembering something is not always a justification for that something. Let��s look at these two premises separately. Premise (1) is supported by the claim that everything that can be remembered must be true and properly grounded, and a belief that is true and properly grounded seems sufficient for knowledge. It is also supported by our reflection into our intuition that we cannot remember something unless we know it. Using Audi��s own example of million-ticket lottery in ��Memorial Justification��. (8) If I have not received news about yesterday��s lottery result, I cannot be said to know the result of the lottery, even if I have a strongly justified and true belief that I would lose. In this case, I cannot reasonably claim that I remember that I lost. Once I get the news for the result and come to know that I have lost, I can retain such belief and be said to remember it. A memorially retained justified true belief does not become a case of remembering until it becomes a case of knowledge. For premise (2), Audi claims that even if a certain piece of belief is vividly remembered, for example I met Max yesterday, I can still fail to be justified in such belief. This is because someone can convince me by good arguments, like citing others�� testimony that Max was in fact not in HK yesterday and it is impossible for me to meet him. In this case, although I in fact had met Max yesterday and it is justified by my vivid remembering, your argument may be good enough to turn down my own justification and it seems that I am not justified in having such belief. Remembering cannot always justify can also be supported by cases of remembering falsely, in the case of memory illusion and memory hallucination. If it is possible for us to remember falsely in these cases, and they are not always distinguishable from our ordinary remembering, it seems remembering itself is not a reliable source of justification. Combining premises (1) and (2), Audi draws the conclusion that it is possible for us to have knowledge that is not justified to be believed.

I agree with the truth of Audi��s premises but I will try to deny the validity of the whole argument. Before doing this, let��s first review once more the process of remembering, and also its relations with beliefs retained in memory and memory beliefs. Using the case of my remembering that Argentina has won the World Cup in 1986. I first had a belief that Argentina had won the World Cup in 1986, and such belief is true and also justified by my watching the TV live broadcast at that time when they won Germany 3:2 in the final. Such belief is then stored in my memory and becomes a retained belief in my memory. Having this retained belief in the mental storehouse of memory available for recall, we can say that the retained belief is remembered. The retained belief becomes a memory belief for me now, because such belief for me on the present moment is grounded on my memory only. I remember that Argentina has won and such belief is true and justified by my memory and it constitutes a piece of knowledge for me now. But suppose I remember falsely that Germany had won, perhaps because I love that team too much so that I have a memorial hallucination for that. My belief that Germany won is a false belief, it is not a belief that can be remembered because it is not true, but it is still a memory belief, a belief that is based on my memory.

Now I will argue that the target being remembered in Audi��s two premises is different. For premise (1), what Audi means by ��remembering something entails knowing something�� is that remembering a given belief X retained in our memory entails knowing X. It is obviously true that my remembering that Argentina won the World Cup in 1990 entails that Argentina really won the World Cup at that year. That Argentina won the World Cup in 1990 is a piece of knowledge for me. If someone convinces me with some genuine argument that in fact Germany had won in that year, then I may not be justified in believing such piece of knowledge anymore. However, what is not justified is not the original belief retained in my memory, but the present memory belief that is grounded on my memory. My present memory belief is not justified because there are convincing counter-arguments and it is also possible for me to remember falsely. But the epistemic defeat of such justification does not pass to the original belief retained in my memory ready for remembering. This is because my original belief retained in my memory does not depend on the same justification for the memory belief. After all, the original belief retained in my memory is still justified by my watching the live broadcast in television. In other words, my original piece of knowledge is still justified, what is not justified is just the memory belief derived from it, which is justified utterly on my memory and remembering of it.

Audi may reply by claiming that the original belief retained and the memory belief so derived is just the same piece of belief, so that when the first one represents a piece of knowledge and the later is not necessarily justified to be believed, it is still a case of having knowledge without justification. My answer is that they cannot be the same because the former represents a piece of knowledge but the latter does not necessarily. If we consider a hypothetical case that my memory belief that Argentina had won the World Cup in 1990 also constitutes a piece of knowledge. Even in this case, such piece of knowledge is utterly different from the knowledge originally we have and stored in our memory. The original belief that Argentina had won is caused by and based on my perception through TV broadcasting, and it represents knowledge even if it is forgotten and fails to be retained in my memory. It is a knowledge for me in the past (i.e. I knew it) and it can exist independently from my present knowledge about it. I knew that Argentina has won, no matter whether I know it now, which in turn depends on whether I can remember it. The memory belief I have now, for Argentina��s victory, is therefore a different piece of belief from my original belief that constitutes my knowledge. We cannot say that the memory belief is just the same as the original belief, just because they have the same proposition or event to be believed.

As a result, Audi is mistaken in claiming that a knowledge entailed by remembering is a piece of knowledge fails to be justified. The piece of knowledge entailed by remembering is constituted by the original belief to be retained in memory, but what remembering fails to justify is a piece of memory belief, which is not yet qualified as a piece of knowledge yet. For the claim that ��knowledge without justification��, if ��knowledge�� means a piece of knowledge in the past, constituted by the original true and justify belief that was retained in memory, then the claim is defeated because this piece of knowledge is still justified at the time that it was formed. Rather, if ��knowledge�� just means the current memory belief remembered, then it is true that remembering fails to justify the belief, but it may be illegitimate to call such belief a knowledge by claiming that it is entailed by remembering. Understanding the relations between remembering, knowledge and memory beliefs, we can see that Audi��s attempt in constructing a knowledge without justification is not a successful one.

(1)��Memorial Justification��, Philosophical Topics 23, 1 (1995), pp. 31-45

(2)Epistemology: A contemporary introduction to the theory of knowledge (New York: Routledge, 1998) Chapter 2 ��Memory��

(3)ibid, p.55,

(4)��Memorial Justification��, p.32

ibid, pp. 70-71

(5) 'a retained memory in belief�� is not a term used by Audi, I articulate this term to denote a particular kind of belief that is different from but closely related to memory beliefs

(6)ibid, p.58, for the sake of my argument, I use retained beliefs instead of facts or propositions for the target of remembering and recalling

(7)ibid, p.56

(8)��Memorial Justification��, pp.41-42

?lt;/P>

previous - next

?lt;/P>