CONTENTS

INTRODUCTION

In the days when I taught philosophy the time available was too short, so my aim was to try to ensure that students learn two main kinds of things.

(a) A family of new concepts that are useful for formulating and discussing philosophical questions.

(b) How to do philosophy. That is very hard to teach. I don't really know how to communicate it except by doing philosophy and helping students to do it, hoping that they will somehow pick up the techniques by imitation and practice. I tried to outline some of the techniques in my 1978 book The Computer Revolution in Philosophy, which is now out of print. However describing the process of philosophising does not seem to be good a way of communicating what it is about to those who have not yet learnt to do it.

To that extent, learning to do philosophy is something like learning to play the violin, or ride a bicycle. It's very easy to play out of tune, or lose your balance.

As this is a single module course you are expected to do on average about 6.5 hours a week of work on the course, including all the time spent in lectures or classes, reading, writing notes, thinking, etc. A significant amount of that time should be spent in the main library or using the internet, finding things out for yourself, e.g. initially by looking for encyclopaedias and dictionaries of philosophy, so that you learn where to find things out quickly, reading books that give overviews, and later on reading recent articles which analyse some of the problems we'll discuss.

You may find it useful to start with an introduction to some fairly 'traditional' ways of doing philosophy. When switching from maths and physics to philosophy as a graduate student, over half a century ago, I found the two books by Bertrand Russell, enormously helpful, even though I did not always agree with him. Or try to find one of the introductory books listed below, e.g. by Campbell, Hospers, Magee (on Popper), Mitchell, or one of the collections of readings in philosophy by Edwards & Pap, Goldman or Lycan. There are very many more books that give introductory overviews, and you will find others by looking in the library, or online, for introductions to philosophy, to metaphysics, to epistemology, to philosophy of mind, to ethics (moral philosophy) or philosophy of science. For the first two weeks of the course, simply try to do as much general reading as you can, and bring to the classes any questions you have about the arguments, or concepts, or theories you find. But before you read anything try thinking about the questions.

The course is assessed by an essay submitted after the course is over. In order to be able to write a good essay you will need practice. So each student will be expected to introduce a discussion of one or two topics this term, and to write a sample essay on which you can get critical feedback. Details will be arranged later.

ASPECTS OF PHILOSOPHY

     “Can a goldfish long for its mother?”
     “If not, why not?”

Note:
There are many variants on the questions about a goldfish. For instance you could ask about other mental states, or about about other animals. Here are some examples (in each case add the the question: “What does this question mean?” “What techniques of enquiry or analysis are relevant to deciding whether one answer is better than another?”

1. Could a mouse desperately hope that her children will do well in life?

2. Could a tadpole hope that it will survive to be a frog and make more tadpoles?

3. Is a fly aware of my fly-swatter coming down to flatten it? (If not, why does it always escape?)

4. Is the fly afraid of being hit by the fly-swatter? (If not, why does if fly away?)

5. Can have robots ability to have hopes, awareness, beliefs, fears, ... How? If not, why not?

These questions relate to several different aspects of philosophy:

Metaphysics: What kinds of things exist? (E.g. fish, material things, mental states, relationships.)

Epistemology (or theory of knowledge): What can we know and how do we know it? In particular how could we tell whether a goldfish does or does not long for its mother? Can we ever know about the contents of “other” minds? For that matter can we know about anything other than our own mind and its contents?

Note: The question “How can we know X?” has at least two very different interpretations. On one interpretation it asks about the causal history of our knowing: did we see the evidence, or hear it, read about it, or infer it from some experiment, use our own tests or belief what someone else said, etc. On the other interpretation it is a question regarding justification: is the means by which you came to know X sufficient justification for claiming that it is true, or that you know it is true?

Philosophy of mind: What are mental states and processes? What's the relationship between mind and body? Are certain material states sufficient to produce mental states? Can mind and matter interact causally? Could longing cause swimming? Can sadness cause weeping? If so, how?

Ethics or moral philosophy: How should we think about the rights of a goldfish? Do we have the right to kill them? To cause them pain? To take them from their mothers? How can such questions be decided? Can ethical questions have right or wrong answers? If so, why is there so much disagreement about them: shouldn't people who disagree be able to find out which side (if any) of the disagreement is right? Compare disagreements about how far the sun is from our planet, or what atoms are composed of.

Philosophy of science: If someone thinks it is a scientific question whether the goldfish feels pain, then we can ask what the difference is between science and other types of knowledge, or knowledge-seeking? What are: scientific theories? Explanations? Evidence? Can theories ever be proved, or refuted, and if so how? What's the relationship between the development of new concepts and the development of new theories?

Conceptual analysis: It soon becomes clear that we are not sure what question we are asking? What does it mean to say that a goldfish longs for something? What does it mean to say that it can think about its mother? Or that it has a mother? (Could a tree or a rock have a mother? What about a battle?) There are many concepts we use outside of doing philosophy, which are extremely difficult to analyse. Examples of such concepts are: mind, matter, meaning, truth, causation, experience, freedom, goodness, concept. knowledge, explanation, science, intelligence, emotion, and many more.

A great deal of modern philosophy attacks old problems by showing that the questions were confused because the concepts used were full of muddles, like “Where is the universe and which way is it moving?”

I believe that learning to do conceptual analysis is one of the most important aspects of learning to do philosophy: everything else hangs on it. E.g. people who are unclear about the concepts they use can argue at cross purposes, or flounder in interminable debates (for examples see most of the discussion in comp.ai.philosophy).

But it is also relevant to being a good scientist. One of the most spectacular examples was how Einstein's attempts to analyse the our concept of simultaneity led to the special theory of relativity.

Last year you had a brief introduction to conceptual analysis when we talked about similarities and differences between: embarrassment, shame, guilt, regret and related concepts. To help you get back into doing philosophy have a go at trying to write down what these states have in common and how they differ.

There is a fairly terse introduction to conceptual analysis in Chapter 4 of The Computer Revolution in Philosophy (1978) now available online here
http://www.cs.bham.ac.uk/research/cogaff/crp/chap4.html

THE RELEVANCE OF AI TO PHILOSOPHY

Equally, AI and Computer Science are relevant to philosophy because they provide a host of new concepts and forms of explanation, as well as raising new questions relevant to old philosophical problems, about metaphysics, about what we can know, about the relationship between mind and matter.

For example: what is a virtual machine? What's the relationship between virtual machines and physical machines? Can virtual machines enter into causal relationships? Can a “software event” like the creation of a new data-structure (e.g. a new list), cause physical events to occur, or is it only physical things that can enter into causal relationships? Do computing systems have “emergent” properties? How do computational machines differ from previous sorts of machines? What are machines? Are connectionist machines significantly different from symbol manipulating machines? What are symbol manipulations? Isn't changing the weight on a neural link a sort of symbol manipulation?

A PROVISIONAL PLAN FOR THE COURSE

(1) Philosophical concepts and jargon. Make the acquaintance of some terminology, e.g. notions like epistemology, metaphysics, ontology, monism, dualism, reductionism, physicalism (materialism), behaviourism, phenomenalism, idealism, epiphenomenalism, interactionism, intentionality, “derivative” intentionality, concept, proposition, rationality, concept empiricism, knowledge empiricism, supervenience, the analytic/synthetic and empirical/apriori distinctions, deductive nomological explanations, the design stance, the intentional stance (Dennett),, and various theories about the relation between mind and body.

(2) Computer science, AI and Philosophy. Try to get a feel for some of the philosophical problems raised by computing and AI. E.g. What is a machine? What is computation? What's the relationship between computational processes and physical processes? Can a software event cause a physical event? What is intelligence? Can there be a behavioural criterion for intelligence? What's the status of the Turing test? Can computational processes support semantics? Are neural processes in some fundamental way different from processes on a digital computer? (E.g. can they discover impossibilities or necessary connections in geometry or topology -- as ancient mathematicians did?)

(3) Machines and intentionality. What does it mean to say that a machine refers to something, or understands something? Searle's Chinese room argument, and replies to it. Causal theories of meaning. Alternatives to causal theories of meaning. Harnad's “Symbol Grounding Problem.”

(4) Representations. What are they? What is their role in intelligence? How many different kinds are there? Does logic have a special role? What, if anything, is special about pictorial or diagrammatic reasoning?

(5) What sorts of machines could have minds? Are there some aspects of mind that are particularly difficult to accommodate in machines in general, or in computers in particular? Qualia? Pains and pleasures? Emotional states? Consciousness? Experience? How could this be settled? Can machines have what Haugeland calls “original intentionality” as opposed to “derivative intentionality”? The relevance of new architectures to new analyses of old concepts. What's the difference between simulation and replication? When is a simulation of a Y a Y?

(6) Freedom of the will and related concepts. What Marvin Minsky called “Dumbell theories” (everything is either an A or a not-A and there's nothing in between) and what is wrong with them. What could it mean for a machine to have its own goals? What does it mean for us to have our own goals? Do we have a kind of freedom machines could never have? Are we (i.e. humans!) not machines?

(7) The importance of architecture. When you have an architecture, it defines a collection of possible states and processes. (Think of how current theories of the architecture of matter define different kinds of stuff - different elements, different chemical compounds, etc. Compare how people previously thought of water, iron, air, etc.) How many of our ordinary mental concepts presuppose specific architectures? What sorts of architectures can support mental states and processes?

Other possible topics to be decided later. Maybe we should talk about consciousness. There are many other concepts of ordinary language that we could try to analyse and relate to the possibility of instantiation in machines, e.g. em sensory experience, learning, desire, pain, pleasure, emotion, personality.

REFERENCES (Always out of date!)

1.BOOKS

    comp.ai.philosophy
    sci.philosophy.meta
    sci.philosophy.tech
    alt.consciousness
    sci.cognitive
    sci.psychology.consciousness
    alt.consciousness
    comp.ai