Memory and Self-Knowledge
Benjamin Winokur (University of Macau)
One difference between human beings and other animals may be our capacity for self-knowledge. The term ‘self-knowledge’ may refer to many things, though. I will talk about just one kind of self-knowledge: the knowledge that you have of your own mental states, such as your beliefs, desires, fears, hopes, tickles, pains, wishes, passing thoughts, imaginings, and more.
These days, many philosophers are quite interested in how you can know your mental states. This is interesting because it seems like this knowledge is quite different from your knowledge of other people’s minds. First, you often seem to know your own mental states better than you know other people’s mental states. Second, you often seem to know your own mental states differently from other people’s mental states.
If you want to know what I want for dinner, you may need to study my behaviour or rely on what I tell you. But if you want to know what you want for dinner, you usually won’t need to put so much effort into finding an answer.
This apparent asymmetry between how you know your mind and how you know other people’s minds has attracted frequent attention from academic philosophers for the past half-century or so, though there is precedent for related discussions in the works of early modern philosophers such as René Descartes or Immanuel Kant.
I will take this occasion at Memory Palace to discuss just one theory of self-knowledge: the so-called transparency theory. I will also narrow our focus, at this point, to self-knowledge of just one kind of mental state: belief. And because you are reading this at the Memory Palace, not the Self-Knowledge Palace, I will try to explain how memory is involved in all this.
Let’s start with the transparency theory of self-knowledge. To introduce the theory, let me share a famous quote from a masterwork of the late, great philosopher Gareth Evans:
If someone asks me “Do you think there is going to be a third world war?”, I must attend, in answering him, to precisely the same outward phenomena as I would attend to if I were answering the question “Will there be a third world war?”.
Evans is describing a procedure for how you know what you believe. He thinks that the procedure is interesting because it is not a procedure for looking inward at your mind. Quite the opposite: using this procedure means that “one’s eyes are directed outward—upon the world”.
It may sound unintuitive that knowing what you believe is achieved by focusing attention away from yourself. You might prefer to say that such self-knowledge is gained through an inward-looking act of “introspection”, whatever this involves exactly. If Evans is right, however, we are wrong to focus on introspection.
The decision to call Evans’ theory of self-knowledge a transparency theory is rooted in metaphor. You want to know what you believe about something, such as whether there will be a third world war. To know this, Evans says that you look to the world beyond your mind.
The question “what do I believe about the world?” is transparent to the question “what is the world like?”. To answer the first question, you answer the second. You look through your belief, so to speak, in order to focus on what your belief is about.
So far, so good…Or is it? Every theory in philosophy faces challenges, and the transparency theory faces several. One is what we might call “the corruption challenge”.
To see the challenge, consider how the transparency theory works one more time. First, you first ask yourself “do I believe X?”. Second, your attention moves to X itself, where X represents some way that the world may or may not be. Third, you judge that X is true, or false. Fourth, and finally, you think “I believe X”, or “I do not believe X”.
The corruption challenge is about what happens at the second step of this procedure. If turning your attention to X causes you to make a fresh judgement about its truth or falsity, then turning your attention to X may cause your mind to change. This will make you unable to report the belief that you have before your mind is changed. And yet we, often want to know what we believe without changing our minds in the process.
Here is where memory might play a role in solving the corruption challenge. The key idea is that turning your attention to X does not always require making a fresh judgement about X. Rather, it sometimes just involves remembering X. When you remember X, you can then say “yep, that’s what I believe!”
What kind of remembering we are talking about here? Following Ben Sorgiovanni, who draws on Endel Tulving’s seminal work in psychology, I suggest that the answer is semantic memory.
Semantic memory enables us to remember information about the world, without experiencing anything about oneself in the process. When you remember a friend’s birthday, or a piece of history trivia, you are not necessarily remembering anything about where you were, or what you were doing, when you learned these bits of information. You are just remembering the information itself. Tulving contrasts semantic memory with episodic memory by noting that episodic memory does involve remembering things about yourself, namely your past experiences.
Now let’s put the concept of semantic memory together with the transparency theory of self-knowledge.
First, semantic memory allows our beliefs about all sorts of things be stored in our minds over time. Indeed, I think that semantic memories are our stored beliefs. Now suppose you ask yourself “do I believe X?”. The transparency theorist says that you answer this question by paying attention to X itself. The new suggestion is that semantically remembering X is one way of paying attention to X.
Now back to the corruption challenge. The challenge is to explain how you can avoid forming a new belief about X when you pay attention to X before answering the question “do I believe X?”. This is because we often want to know what we believe without changing our minds in the process. The solution is to use your semantic memory.
By remembering X, you do not need to make a fresh judgement about X. You just bring your pre-existing belief about X into mind by remembering X.A problem with this new argument is that, when “X” pops into your head after asking yourself “do I believe X?”, it can be difficult to know whether you are remembering X as opposed to merely imagining X or thinking about X in any other non-believing way.
It is not easy to solve this new problem as a transparency theorist. This is because distinguishing your semantic memories—hence your stored beliefs—about X from other ways of thinking about X seems to require paying attention to the different features of your different mental states, and this sounds like an inward-looking task.
What should the transparency theorist do about this? I have thought of a possible solution, but it admittedly has serious limitations.
My solution is to say that, when the thought “X” pops into your head after asking yourself “do I believe X?” you are entitled to treat “X” as a manifestation of your belief—a semantically remembered belief—that X is true.
So, when you ask yourself “do I believe X?” and the thought “X” pops into your head, I suggest that you are entitled to treat “X” as an expression of your belief because you are the kind of being—a human being—that is especially dependent on semantic memory for holding your beliefs in mind over time. You do not rely on mere passing thoughts about X, or imaginings about X, for any equally important cognitive purposes. So, you can ignore the possibility that “X” merely manifests a non-believing way of thinking about X.
Maybe you like this approach. I do too, to some degree. The problem is that it only explains why you are entitled to treatX as what you believe. In other words, it does not explain how you can ever know that you are right. What if you are entitled to treat X as what you believe and yet you are actually incorrect? In this case, you are self-ignorant, even if you are also entitled to your self-ignorance.
At present, I have not seen a better way to explain how memory connects with self-knowledge. Maybe there is a way, or maybe there is not. Either way, the issue is rather underexplored, though perhaps things will change soon
