Aaron Sloman
School of Computer Science
The University of Birmingham

Viewable in ANY browser.

There's a file pointing to recent research papers and discussion notes.

The Cognition and Affect Directory has many papers listed, with abstracts, in the index.

My "miscellaneous" directory has a disparate collection of draft papers, discussion notes, email discussions, news postings, written at various times.


I use the Poplog development environment for my teaching and my research. Over the years my colleagues and I have developed many teaching tools including software and online documentation introducing topics in AI, Cognitive Science, Linguistics and Computer Science through a range of practical programming-based activities. Most of the material is available in the Birmingham Free Poplog directory especially in the teach/ subdirectory.


Among other things I have been working on several papers concerned with the nature of mind and its relation to matter, and ways of generalising computer science to include AI, theoretical biology, the study of human development, and comparative cognitive science.

Design spaces and niche spaces

This draft, incomplete discussion note can be found in my "miscellaneous" directory in the plain text file entitled designs.and.niches

The evolution of what?

A draft (still evolving) paper on the evolution of consciousness is available in compressed and uncompressed postscript format in the Birmingham Cognition and Affect project directory:

in these two files

Supervenience and Implementation

An evolving paper on the philosopher's notion of supervenience and its relationship to the engineer's notion of implementation can be found in my "miscellaneous" directory in the plain text file called supervenience.

Software development

I am also continuing work on the SIM_AGENT toolkit located in the Free Poplog ftp directory also accessible from and described in this Web page

I have built, with help and advice from colleagues and student testers(!), an object oriented collection of graphical tools required for the toolkit and also usable for other purposes in connection with Pop-11. These are browsable in the Poplog directory in the rclib sub-directory: and can be fetched as gzipped tar file


Actual Possibilities

A very general exploratory paper presented at KR96 at MIT in November 1996, and included in the published conference proceedings.

The paper tries to tease out some of the general ontological assumptions lying behind naive (and sophisticated) physics and issues about design and causality. Treats objects as "possibilities transducers", so that a complex design is a network of possibility transduction links (including feedback). It claims that there is no inherent difference between physical causation and causation in virtual machines, which I later hope to expand into a defence of functionalist analysis of mind (consciousness, etc.) No formulas yet, but ultimately needs some -- perhaps a new alternative to modal logic?

I think this links up with some of Judea Pearl's work, but I don't know enough maths to follow all the details. I may also have under-rated the work of modal logicians.

There are also links to Lotfi Zadeh's recent work on fuzzy granularity. I hope to spell out these links at some later date.

Both papers are available in postscript format if requested. The second paper is relevant to an informal discussion paper containing some notes on indeterminacy in classical physics partly arising from correspondence with Henry Stapp on quantum physics and consciousness.

I now feel that classical physicists never could have assumed that objects had infinitely precise positions, sizes, locations, etc. Therefore many physical devices (with non-linearities) can be seen as amplifiers of inherent indeterminacy, and in some cases (e.g. gambling machines) they use geometry to force indeterministic systems to select between determinate outcomes. On this view classical physics assumed deterministic equations but not determinate initial conditions (inputs to the equations). I.e. classical physics thus construed has some of the characteristics of quantum physics that are a cause of much debate and confusion.

But I don't know enough about either classical or quantum physics, so any help will be much appreciated. The relevance to computing is related to the need to understand various kinds of machines and the extent to which they are or are not deterministic and how certain designs can impose structure on the outcomes of partly indeterministic processes.

It also relates to my long term goal of clearing up muddles about consciousness, qualia, etc. (I still feel that a huge amount what's written about such matters is essentially rubbish, as I am sure will become clear some time in the next hundred years or so.)

This (lynx friendly) file is maintained by

Aaron Sloman.

Last updated 1 May 2000