cognitive science and philosophy of mind.
And if the title is not entirely clear,
I will also read it in Dutch... "Cognitie, mechanisme en betekenis:
pleidooi voor het socionormatieve binnen cognitieve wetenschap en
filosofie van de geest." Because the PhD candidate - for he is officially still a candidate - has met all the requirements of the PhD program of the University of Antwerp, he is allowed
to publicly defend his dissertation, which he shall now do. I'm very pleased to give the floor to Farid Zahnoun. He will give a summary of his dissertation of about twenty minutes, and he will do so in Dutch.
After that, we will switch to English. Farid, the floor is yours. Thank you. Yes, well, twenty minutes is a little short to explain what I've been trying to achieve in the past four years.
And actually, I've been trying to achieve a number of different things. To be honest, it seems a bit pointless to explain in twenty minutes what I've been working on these past four years, not because things would become too technical, but simply because the relevance of my work is impossible to understand without having at least some knowledge about the general philosophical background against which my dissertation should be read. Therefore, I felt it would be more interesting to explain to you today in those twenty minutes, or maybe a bit longer, what this broader philosophical background is.
Now, I'm aware that some of the things I will say might be a bit difficult to understand, but rest assured, after my lecture comes the actual defense which will be even harder to follow. So the point of this presentation is that afterwards the audience has at least some idea of what the title of my dissertation means.
I will first say something about my background. The kind of philosophy I'm interested in is called 'philosophy of cognitive science'. Now, what is cognitive science? Maybe some people here have heard of the term, although it might be a bit misleading. Cognitive science is not a typical branch of the sciences, like physics, biology or chemistry.
Rather, it is an interdisciplinary enterprise with contributions coming from psychology, neuroscience, computer science, linguistics, philosophy, and also from robotics and research in the field of
artificial intelligence. So we might wonder, what keeps cognitive science together?
Well, first of all a common goal, which is to reach a better understanding of animal and, especially, human cognition.
Now, cognition is itself a vague term but for simplicity's sake we could say it refers to those things we ordinarily call 'mental', including things like perception, memory, language acquisition, learning processes, thought processes and so on. And secondly, cognitive science is also kept together by a number of shared assumptions. One of these assumptions takes center stage in my research, namely, the assumption that all forms of cognition need to be explained by looking at the brain and that the brain is, in essence, a kind of information-processor.
It is actually useful here to also give you a bit of historical background. Cognitive science is actually not that old and, in fact, the term itself was only coined in the seventies. But the first developments towards cognitive science took place in the fifties. Two of these developments deserve special mention.
First, there was the growing discontent with behaviorism. In the fifties, more and more people started to question the behavioristic approach, which wanted to reduce the study of the mental to a study of behavior (hence the term). So more and more people felt this approach to be lacking. And, second, there was the development of the computer.
These two developments somehow come together, leading up to the fundamental idea that the brain is a kind of biological, digital computer that performs computations over internal symbols or representations. I will clarify this in a minute. But the idea of the brain as an information processor didn't just drop out of the sky. The idea made sense because it is affiliated with a much older, more general philosophical idea, namely representationalism.
Representationalism claims that brains are first of all in the business of representing the outside world. Of course, people have always known that our bodies contain all kinds of organs which all have some function to perform, and for most of these organs, we already know long since what these functions are. Except for the brain. It is pretty hard to say what brains do.
Well, the longstanding assumption is that brains are mainly serving the purpose of representing the outside world to us.
So this notion of representation becomes crucial.
Now, the idea that brains represent the world and that the body is only of secondary importance is actually quite popular. Take, for instance, the brain-in-a-vat hypothesis... Perhaps the hypothesis itself is not that popular, but at the same time, the entire film 'The Matrix' is built on this premisse.
So it is in principle possible that we are just brains in a vat, hooked up to a machine that provides stimuli which generates as its output a reality which is in reality not there. So the brain does all the work.
For people who are to lazy to go for a run (see slides), this would be convenient. So the main question now becomes: how does the brain manage to do this? This brings us to our central notion, which is the notion of internal representation. So the idea is that the brain somehow manages to construct internal representations, or a kind of copies, and then performs computations over these entities. Now, this notion of representation, which was already crucial in the early stages of cognitive science is today still the most important explanatory notion within cognitive science.
Now, what are internal representations exactly? After sixty years, the situation is still that almost everyone agrees that we have to assume the existence of representations, but at the same time nobody knows exactly what they are or how they are being realized in the brain. So the situation is quite paradoxical. But there is nevertheless some minimal agreement on what these representations are supposed to be doing. So, representations are 'things' that somehow say how things are in the world and which , in that sense, represent these things.
So this also means that they have a kind of content, in the way sentences or or ordinary public representations (I'll give examples) are said to have content. These are things that 'are about' other things and they can have semantic content in the way a sentence can have such content. And these inner representations are indeed modeled on these ordinary representations like sentences, but also pictures, models, maps, in short, al those things we think about when we think about representation. And my claim is that the idea of internal representation is untenable.
This is the main subject of my dissertation. Here, I want to give arguments for this claim, not based on what is literally in my dissertation - this would take us too far - but I realized I can easily clarify my claim by making use of a very interesting distinction which the American philosopher Wilfrid Sellars made in the sixties, namely the distinction between the manifest and the scientific image. In a very famous article, Sellars begins by presenting us with an interesting definition of philosophy. "The aim of philosophy, abstractly formulated, is to understand how things in the broadest possible sense of the term hang together in the broadest possible sense of the term.
I've also translated it here. This is an oft-quoted definition. What Sellars is saying here is that the philosopher is in this regard confronted with a very special problem. There is actually not one, but two ways of having an image of the world and, in a certain sense, both images are correct.
And he captures these two different images in terms of the manifest and the scientific image. So I will first say something about the manifest image, and then move on to the scientific image.
The manifest image is probably easiest to understand. It refers to our daily reality in which objects and events appear, not so much as what they are in and of themselves, but first of all as what they for us, humans. In other words, they appear in relation to who we are and what we do.
In a sense, then, from this perspective, the world already has meaning. All of this is, of course, very similar to what Heidegger calls "world'. So, take as an example, almost everything you can see around you in this environment: the chairs you sit on, the books here in the library, the library itself, or, for instance, money... I wanted to take some out of my wallet as an illustration, but unfortunately I spent everything on the reception from later.
But, of course, also human behavior is like this. So, a chair is only a chair in relation to us. In itself, there are no chairs. And there certainly are no books in itself, let alone libraries.
It is only because we use these objects within certain practices that these things first acquire their identity. And we also see them first of all as what they are for us. Now, the term 'manifest', in relation to 'manifest image', should be understood in terms of how these things manifest themselves to us, and indeed, how we perceive them quite literally. And, as said, the same goes for human behavior.
Human behavior is is of course crucial here since we are talking about psychology. And Sellars claims that, from the perspective of the manifest image, we see human behavior first of all as following from reasons. People act for a reason, which is meaningful. They don't act from causes.
So if we want to know why a person does something, we are not interested in certain causal mechanisms or in how the brain makes the person move. Ordinarily, we try to come up with an answer in terms of what the person wants to do, so we look for a reason. And reasons are meaningful. All of this stands in sharp contrast to the scientific image.
From the perspective of the scientific image, reality is being approached as it is objectively, which means, how it is, precisely NOT for us. We want to know what things are, in separation from our own interests or our daily ways of dealing with these things. The scientific image, then, is first of all interested in causal relations, not in relations of meaning. So this is also why Sellars mentions the idea of 'hanging together'.
So actually, there are two ways in which things, as a whole, hang together, but these are two totally different ways. All of this can be clarified by the use of some examples. Take, for instance, gold. Before, I gave the example of money.
But now consider gold. To us, one kg of gold is something which has the value of about 36.000 Euro. From a scientific perspective, however, gold is only matter whose physical and chemical . Properties are pretty well known to us.
Here are a few of these properties (on slide)... Things like density, tensile strength, ...I also don't know what these things mean exactly...Another, perhaps more familiar example, is sound versus music. So, from the point of view of the scientific image, suppose someone asks you: "What am I hearing?" Then one can reply: "This is a piano sonata by Mozart." But the other person might also mean that he wants to know what that music really is, scientifically speaking. Well, music is sound, and sound is nothing but our perception of changes in air density.
Here's the definition of sound we find on Wikipedia. I didn't make up these things... And then a third example is that of a geographical map. This is actually a useful example for my purposes because a map is already an example of an ordinary public representation.
So, a map has physical properties, and has indeed a structural similarity with the terrain it maps, but crucially, the structural similarities do not themselves explain how the map manages to represent or, in this example, that this dot represents Sydney Oprah House, and the pink line next to it the Harbour Bridge. Indeed, I chose this example especially for Dan Hutto. I first wanted to use a map of Wollongong, but probably no one here has heard of the place....Sorry... So the central problem now becomes, and this is the question the representationalist cognitive scientist must answer: "What makes something a representation?" If it's not structural similarity, then what is it that makes it the case that some object can stand for something or that some object can refer to something? Well, my insight - which is actually not that spectacular - is that the representational aspect of representations is impossible to understand from within the scientific image.
To be clear, this is continuously being attempted by philosophers who try to give a kind of objective foundation to the fact that something represents in terms of the properties of the object. So, returning to the scientific image, of course scientists not only want to describe reality, they also want to explain it. In this regard, Sellars has another interesting thing to say. He claims that what makes the scientific image unique is, not simply that it wants to explain, for this also happens in the manifest image.
In our daily lives, we also try to explain things. But what makes the scientific image so unique lies in the specific way in which it tries to explain things. What, then, is this unique way of explanation? According to Sellars, the scientific image is characterized - and I will read it in Dutch - by the fact that it wants to explain manifest phenomena by postulating imperceptible, and hence non-manifest, entities. So scientific theories postulate things 'behind' the manifest phenomena, on the basis of which these phenomena can be explained.
And, indeed, on closer inspection, cognitive science also does this. But I will first give some examples. When we want to know in the manifest image why the lights are on here, then it suffices to mention that someone flipped the light switch. What we won't do is give an explanation in terms of the movements of electrons, which is precisely what the scientist does want to do.
If you really want to understand why the lights are on you have to refer to electrons, which are, of course, imperceptible entities, like atoms, black holes, gravity, and also germs.
Of course, today we know that viruses exist, but already in 1546, a theory was proposed by Girolamo Fracastoro, who claimed that diseases are transmitted through invisible entities. Unfortunately, it would still take a few centuries before the theory caught on. And also within cognitive science, we find such imperceptible entities, namely these representations in our heads. Now, as just said, they are imperceptible as well as causally explanatory, yet - and this sets them apart from other postulated entities- internal representations are about things in the world.
They somehow say how things stand with the world. So again, the question is, how do they do this? And again, I say that this question is unanswerable from within the scientific image. But it is answerable from within the manifest image. Within the manifest image namely, things have meaning.
This defines our primary and fundamental way of relating to the world. Things have meaning. We are not first of all interested in what these things are in and of themselves, but first and foremost in what they already actually mean, or might potentially mean. And, of course, I'm using 'meaning' here in a very wide sense.
We could also put it by saying that the world is first of all related to in its actual or potential significance. Now, people also express this meaning and communicate it to others through language. Humans have this special ability of representing this meaning by using language. Now we're talking about meaning in a very specific, semantic sense.
But this means that, with language, a whole new set of meaningful objects enters the manifest image, namely things that are about other things. I am first of all here thinking of spoken language and, coming later, written language. These objects are characterized by what philosophy calls 'intentionality'. 'Intentionality' is a very important notion, especially in philosophy of consciousness because intentionality is said to be the defining property of consciousness.
Consciousness is always 'of' something. It is in some way always about something. So consciousness isn't just a thing among other things, it is always 'of' something. There is of course discussion about whether this is true, but this is at least how consciousness is put forward since the late 19th century.
Now, mental states like beliefs, ideas and desires have this property. They are about something. But also certain external objects seem to have this property of being about something. So, one common example of this, which keeps returning in handbooks, is the sentence "The cat is on the mat".
From a scientific perspective, there are just some black spots on a white screen in a certain order, but for us, these things have meaning. They are about something. They are about a cat, although we don't really know which cat. And apparently, she is sitting on a mat.
And in addition, this can be true or false. So this is also an important feature of statements. They can be true or false. Of course, things in themselves can't be true or false.
It is simply nonsense to say that "the cat is false". This makes absolutely no sense. "The mat is true" is equally absurd. However, the sounds I'm producing when I say "The cat is on the mat", can nevertheless be true or false.
I have still another example of a sentence, written by the Flemish author Koen Sels, I suggest we now read the sentence in silence...(See 'Generator' by Koen Sels, p. 13) No, I'm not going to read it out loud. The sentence is too long, but I think it is one of the most beautiful lines I have ever read. The author describes a childhood experience of having to celebrate Christmas in the house of his grandparents, somewhere in a small Flemish hamlet.
In case you want to read the whole sentence again, I suggest you buy the book. On this picture, one sees the author, shortly after his book presentation. Anyway, sentences are about something. But not only sentences, also other objects.
Take, for instance, this trivial weather icon. I give this example to emphasize that things we can use to infer information do not necessarily represent anything. Suppose you want to know if it's going to rain. You could just look at the clouds.
But these clouds do not represent the fact that it is going to rain. They do not stand for anything, they just are. We can use them as a sign of the coming rain, but this doesn't make the clouds representations. On the other hand, if you want to know whether it's going to rain, you can also enter 'weather' in Google, and you'll see such a weather icon appear.
This icon does represent. It really stands for the fact that it is going to rain. However, in itself, the icon is just another physical object. Maybe it resembles a rain cloud, but this is not a necessary condition.
Sentences don't resemble what they are about. So once again, the question is, what makes it the case that such an object can represent something, in this case a rain prediction? And my answer is, this question is already poorly formulated. It is not a 'what' that makes an object a representation, but a 'who'. Here is my general definition of representation: a physical object is a representation with a specific semantically evaluable content if and only if it is to be regarded as such within a community.
And here we see the element of social normativity. So representation is inherently a socionormative notion. I'll clarify this a bit, for this might sound a bit counterintuitive. So the definition contains two essential components, first, representation presupposes the cognitive capacity to see something as standing for something else, or to see something as something it is not, but which is absent.
This is in itself already a very complex cognitive capacity. Humans have this ability, but I doubt whether animals can do this too. Secondly, representation presupposes that this capacity is also socionormatively regulated. So someting can only say something about something if it can do this in principle for a community.
Returning to the example of the weather icon, the fact that it stands for the prediction that it is going to rain is itself the consequence of the fact that not just I, but we, are expected to see that object in this way. This means that a normative element, in this case a kind of social imperative, which is of course learned, generates an objective reality. In other words, here, a prescriptive reality constitutes a descriptive reality. And this is something which cannot be understood from the scientific image.
So, nothing represents in and of itself, outside our soconormative practices in which we first of all learn, and no so much discover, in the way the scientist might discover that an object has certain properties, that an object says something about the world, as well as what it says about the world. So, and I have reached the final slide, what does this mean for cognitive science's notion of internal representation? Well, it should be clear that representation belongs within Sellars' manifest image and not at all in the scientific image. But, of course, there is a good reason for why cognitive science wants to hold on to the notion of representation. The point is to use a notion which does explanatory work within the manifest image, in general, meaning, or in fact anything involving semantic content, and to use this within causal explanations.
So something from the manifest image is smuggled into the scientific image. And this is, of course, not done. Moreover, as said, representation presupposes complex cognitive capacities. It is precisely the job of the cognitive scientist to explain these capacities, not to use them as explanatory posits.
If there is anything that needs explaining, it is precisely representation, or language, or the fact that we experience the world as meaningful. So what we see is a certain reversal of explanandum and explanans. So, internal representation as a bridge between manifest and scientific image? (Here's the Harbour Bridge again) No, rather a notion that falls in between the cracks and disappears in the gap. Ah, I thought this was the least funny bit...Thank you..
No comments:
Post a Comment