From newshub.ccs.yorku.ca!torn!cs.utexas.edu!zaphod.mps.ohio-state.edu!caen!deccrl!bloom-beacon!eru.mt.luth.se!lunic!sunic!seunet!his!mikael Wed Aug 12 16:52:16 EDT 1992
Article 6556 of comp.ai.philosophy:
Xref: newshub.ccs.yorku.ca comp.ai.neural-nets:3840 sci.cognitive:281 comp.ai:3822 comp.ai.philosophy:6556 sci.philosophy.tech:3379
Path: newshub.ccs.yorku.ca!torn!cs.utexas.edu!zaphod.mps.ohio-state.edu!caen!deccrl!bloom-beacon!eru.mt.luth.se!lunic!sunic!seunet!his!mikael
>From: mikael@ipcsun1.his.se (Mikael Boden)
Newsgroups: comp.ai.neural-nets,sci.cognitive,comp.ai,comp.ai.philosophy,sci.philosophy.tech
Subject: SNCC-92 Conference programme
Message-ID: <MIKAEL.92Aug4105544@ipcsun1.his.se>
Date: 4 Aug 92 08:55:44 GMT
Sender: news@his.UUCP
Followup-To: comp.ai.neural-nets
Distribution: comp
Organization: University of Skoevde
Lines: 661



              The Connectionist Research Group
                    University of Skovde


                         The First
              Swedish National Conference on
                       Connectionism
 Wednesday 9th and Thursday 10th Sept. 1992, Skovde, Sweden
                             at
          Billingehus Hotel and Conference Centre


                         PROGRAMME

                        Secretariat:
                          SNCC-92
                    Attn: Ulrica Carlbom
                    University of Skovde
                        P.O. Box 408
                  S-541 28 Skovde, SWEDEN
        Phone +46 (0)500-77600, Fax +46 (0)500-16325
                     conference@his.se

                   Conference organisers
     Lars Niklasson (University of Skovde) lars@his.se
     Mikael Boden (University of Skovde) mikael@his.se

                     Program committee
   Anders Lansner (Royal Institute of Technology, Sweden)
         Noel E. Sharkey (University of Exeter, UK)
         Ajit Narayanan (University of Exeter, UK)

                    Conference sponsors
                    University of Skovde
  The County of Skaraborg (Lansstyrelsen, Skaraborgs Lan)

                     Conference patrons
  Lars-Erik Johansson, Vice-chancellor University of Skovde
 Stig Emanuelsson, Head of Comp. Sci. Dept., Univ. of Skovde


                 Background and objectives

During the 1980s, Connectionism was established as  a  major
computational paradigm. It is distinguished from traditional
symbolic approaches by its use of dynamical systems describ-
able  by  mathematical equations. It offers a radically dif-
ferent conception of the basic operations of the  mind-brain
by  appealing,  not to classical symbol manipulation, but to
causal processes by which  units  excite  and  inhibit  each
other.

The conference is intended to put the  Swedish  research  in
perspective with the international research and the state of
the art of connectionism and neural networks. It  is  impor-
tant  to  be aware of what research (both basic research and
state of the art) is carried out in this area, both  nation-
ally  and  internationally.  The  conference will serve as a
forum for sharing ideas and knowledge in connectionism.

The conference will provide a forum for technical and philo-
sophical exchange between all members of the world-wide con-
nectionist community.  There will also  be  a  philosophical
stream  during  the conference, in which the implications of
the latest connectionist research for the mind/brain problem
will be discussed.

The Swedish Neural Network Society (SNNS) will hold an offi-
cial members meeting at the conference.


                Abstracts of Invited Papers

Connectionism and the Mind-Body Problem: Exposing  the  Rift
between Mind and Cognition.
Dr. Tim van Gelder, Assistant Professor, Department of  Phi-
losophy  and  Cognitive  Science Program, at Indiana Univer-
sity, USA

   Insofar as there is an orthodox view on  the  ontological
   mind-body  problem in contemporary philosophy, it is that
   mental states and  processes  (as  they  figure  in  folk
   psychology)  are  token-identical  with  brain states and
   processes. One such theory is computational functionalism
   of  the  "Language of Thought" variety advanced by Fodor,
   Pylyshyn and others. As a number of authors  have  noted,
   however,  some kinds of connectionist models of cognition
   appear to cast doubt on token identity theses since there
   is  nothing  in them that the entities of folk psychology
   might be plausibly be identified with. These models  thus
   seem to support a form of eliminativism about the mental.
   I will argue that this kind of  eliminativist  conclusion
   is  mistaken.  Both  eliminativism  and standard identity
   theories are founded on the same deep  Cartesian  miscon-
   ception  about  the mind, namely, that mind is inside the
   head and causally  responsible  for  our  behavior.  This
   misconception  has  already been diagnosed by people such
   as Ryle and Heidegger. The upshot  of  connectionism  for
   the  mind-body problem is that we should supplement their
   perspectives by distinguishing between mind, on one hand,
   and  cognition,  on the other. Mind and cognition are not
   the same thing, even though some cognitive  processes  do
   fall  under  the traditional umbrella of mind. It follows
   that connectionism and  cognitive  science  are  not,  in
   fact,  the study of the mind as such - they are the study
   of cognition, which is just one aspect of mind. If  space
   allows,  I will discuss a further upshot of connectionism
   for  the  mind-body  problem,  Smolensky's  very  general
   empirical  thesis that cognition is state-space evolution
   in high-dimensional non-linear dynamical systems.


Explaining Cognition with Nonlinear Dynamics.
Dr. Jordan B. Pollack, Assistant Professor Laboratory for AI
Research Ohio State University, USA

   While it is generally agreed that  it  is  the  nonlinear
   nature of modern abstract neural networks which give them
   more power than single layered linear  systems,  much  of
   the  80's  research  has  focused  on  surpression of the
   unstable, unpredictable, and unanalysable nature of  non-
   linearity  through  simplifying  devices  such  as quasi-
   linear functions or symmetric weights. Chaos, in the form
   of     pseudo-randomness    or    sensitivity-to-initial-
   conditions, is usually considered a "bug." However, there
   are  modes  of  complex non-linear systems which have the
   necessary "features" to provide powerful alternatives  to
   symbolic  theories,  and I will relate these mathematical
   properties of neural  networks  to  generative  cognitive
   faculties such as language and imagination.



Semantic and Syntactic Decompositions of  Fully  Distributed
Representations
Dr. Noel E. Sharkey, Centre for Connection Science,  Depart-
ment of Computer Science,University of Exeter, U.K.

   A common view of the role  of  connectionist  models  for
   cognition  is  that  they  should only be used to perform
   operations requiring fast parallel  memory  access  while
   manipulative operations, e.g. those sensitive to the syn-
   tactic structure of representations,  should  be  carried
   out  by  a  symbolic  mechanism. This is in line with the
   Classical view  of  cognition  (e.g.  Fodor  &  Pylyshyn,
   1988).  The  implication  being  that  connectionist com-
   ponents act as subroutines that must emerge  onto  symbol
   surface to return combinatorial expressions that are con-
   sonant with a Classical  system.  i.e.  expressions  with
   symbol  tokens that maintain the structure of expressions
   in position vectors or their  equivalent  (cf.   Sharkey,
   1991  for  a review). Moreover, Fodor & McLaughlin (1990)
   argue that "...  whereas the Classical constituents of  a
   complex  symbol  are, ipso facto, available to contribute
   to the causal consequences of its tokenings - in particu-
   lar,  they  are  available  to provide domains for mental
   processes - the components of tensor product and superpo-
   sition  vectors  can  have no causal status as such." (p.
   198)

   Counter evidence to these  critiques  will  be  presented
   here  in  four  parts.  First, a broad notion of composi-
   tionality will be discussed, followed by a  short  review
   of  some  recent  computational  results on the structure
   sensitivity of superpositional  representations.  Second,
   an  analysis  will be given of how contextually invariant
   weight representations are composed onto hidden units  in
   a  way that is contextually sensitive. Third, a method is
   described for the decomposition of the representation  of
   constituents  from  complex  superpositional expressions.
   This will be compared with a method proposed  by  Smolen-
   sky.  Fourth, two computational studies will be presented
   which use the method of decomposition to demonstrate  the
   causal  efficacy  of  the  constituent representations in
   both a structural  disambiguation  task  and  a  semantic
   disambiguation task.


SCRuFFy:       An        applications-oriented        hybrid
connectionist/symbolic shell.
Dr. James A. Hendler, Associate Professor Dept. of  Computer
Science UMCP, USA

   There is a clear and present need for systems  which  can
   integrate  perceptual  components,  currently best imple-
   mented by connectionist  models,  and  expert  reasoning,
   currently  being best solved by AI systems. To facilitate
   the near-term development of such systems  we  have  been
   working  on  the  development of an applications-oriented
   shell providing an integration of  a  connectionist  back
   propagation   learning  network  with  an  expert  system
   development language.  We have developed the SCRuFFy sys-
   tem  as a prototype shell for the development of applica-
   tions integrating neural network and expert systems  con-
   trol.  The  system  provides for the integration of these
   components in a blackboard-based  architecture,  and  has
   been  tested in several areas. We will discuss the design
   of the shell and show how it has been  used  for  several
   different  domains  in  which sensor-based information is
   used as the basis for intelligent decision making.


Neurocomputing Hardware: Present and Future.
Dr. Dan Hammerstrom Adaptive Solutions, Inc., USA

   This presentation will discuss the current state  of  the
   art  of  industrial neurocomputing, and then speculate on
   its future.  Three examples of  commercial  neuro-silicon
   will be presented: the Adaptive Solutions's CNAPS system,
   the Intel ETANN chip, and the Synaptics OCR  chip.  These
   will be discussed in the context of a review of the basic
   issues involved in neural network hardware  such  as  the
   sum  of  products computation, weight storage, and inter-
   processing element signal communication.  We then  specu-
   late  on  where  commercial  neurocomputing  hardware  is
   going. In particular, we propose that commercial  systems
   will  evolve  in  the direction of capturing more contex-
   tual, knowledge level information.  Some  results  of  an
   industrial   handwritten   character  recognition  system
   created at Apple Computers will be presented which demon-
   strates  the  power  of  adding  contextual  knowledge to
   neural network based recognition. Also discussed will  be
   some  of the possible directions required for neural net-
   work algorithms needed  to  capture  such  knowledge  and
   utilize  it effectively. Some results from experiments on
   capturing contextual knowledge  using  several  different
   neural  network  algorithms  will be presented.  Finally,
   the issues involved in designing VLSI  architectures  for
   the  efficient  emulation of sparsely activated, sparsely
   connected contextual networks will be  discussed.  It  is
   shown  that there are fundamental cost/performance limits
   when emulating such sparse structures in both the digital
   and  analog  domain.   This is joint work with Steve Reh-
   fuss, Department of Computer Science/Engineering,  Oregon
   Graduate Institute, USA


Structure and Change in Connectionist Models.
Dr. Jerome A. Feldman, International Computer Science Inst.,
University of California, Berkeley, USA

   For a variety of reasons, connectionist or neural network
   research is often identified with the weight modification
   techniques. An extreme form of this  view  considers  any
   designed structure as violating connectionist principles.
   This view would be fatal to the enterprise because intel-
   ligence  requires  elaborate innate structures as well as
   adaptation. The talk will present a current  overview  of
   structured   connectionist   modeling   and  some  hybrid
   efforts. One focus will be the task of  learning  natural
   language from examples.


On Nativist Connectionism.
Dr. Ajit Narayanan Dept. of Computer Science  University  of
Exeter, UK

   Ramsey and Stich  have  assessed  connectionist  language
   models  with respect to nativist claims that human beings
   have a rich store of innate knowledge. The `argument from
   the  poverty  of the stimulus' is claimed by nativists to
   demonstrate that empiricist theories of the mind are mis-
   taken.  Ramsey and Stich point out that connectionism can
   challenge nativism not only because  it  uses  empiricist
   techniques  but  also  because  it rejects the idea of an
   internalized grammar. Ramsey and  Stich's  assessment  is
   that  the poverty of the stimulus argument, and hence the
   argument for nativism, can be reconstructed, despite  the
   connectionist  challenge. However, it will be argued that
   Ramsey and Stich have misidentified  the  nature  of  the
   relationship  between  connectionism  and  nativism. This
   they may have done by, first, regarding the  terms  `con-
   nectionism'  and  `empiricism' as largely synonymous, and
   secondly, by assuming that the only nativist  is  a  sym-
   bolic or generative grammar nativist adopting a deductive
   approach.  The main aim of this paper is to explore three
   possible  connectionist  stances  on innateness. We shall
   argue that connectionists can be nativists without neces-
   sarily  being,  for  want  of  a  better  term, `symbolic
   nativists'. Also, connectionist versions of  the  innate-
   ness hypothesis can be just as strong in their commitment
   to nativism as symbolic versions. What  is  different  is
   what it is that connectionists commit themselves to.

The neuropharmacology of associative memory function: an  in
vitro,  in  vivo, and in computo study of object recognition
in olfactory cortex.
Dr. James M. Bower, Associate  Professor  Div.  of  Biology,
California Institute of Technology, USA

   For the last several years  we  have  been  studying  the
   organization  of  olfactory  cerebral  cortical  circuits
   using a combination of experimental  and  modeling  tech-
   niques.  Much  of our early effort involved the construc-
   tion of a realistic numerical simulation of the  piriform
   cortex  capable of replicating complex dynamical patterns
   of network activity.  Using  this  model,  we  have  more
   recently  been  exploring  the ways in which this modeled
   network could recognize  complex  patterns  of  olfactory
   stimulation  through  the biological implementation of an
   associative memory. Through the  ongoing  interaction  of
   modeling and experiment investigations we have discovered
   a potentially significant  role  for  the  neuromodulator
   acetylcholine  in  the memory storage and recall process.
   If correct, these results have important implications for
   both  the  function  and disfunction of cerebral cortical
   circuits.


Connectionist Cognitive Maps and the Development  of  Objec-
tivity.
Dr. Ronald L. Chrisley University of Sussex, UK

   Cognitive science should not  concern  itself  only  with
   explaining  mental  phenomena that represent the world as
   an  objective,  subject-independent  realm  filled   with
   objects  and properties that are logically independent of
   each other. Cognitive science should also explain  inten-
   tional phenomena that are pre-objective, especially those
   that permit  the  construction  and  maintenance  of  the
   framework  which makes objective representation possible.
   Therefore, insofar as classical architecture is the para-
   digmatic  ground  for objective, conceptual cognition, it
   cannot  be  a  ground  for  these   pre-objective,   pre-
   conceptual abilities.  Three questions immediately arise:
   (1) How can we explain pre-conceptual phenomena? What are
   some  of  the  philosophical  difficulties  this  form of
   explanation generates? What reason do we have to  believe
   that  such  phenomena are representational, and therefore
   the proper subject of  intentional  analysis?   (2)  What
   kind  of  pre-conceptual abilities can result in the con-
   struction and maintenance of conceptual  abilities?   (3)
   What  kind of cognitive architecture *is* appropriate for
   pre-conceptual cognition?  After (briefly) answering  the
   questions  in  (1),  and  (briefly)  suggesting  that the
   notion of a cognitive map might be  the  answer  to  (2),
   I'll  focus  on  (3).  I'll  provide  several reasons for
   thinking the idea that connectionism is  well-suited  for
   grounding  pre-conceptual  intentionality.  With all that
   by means of theoretical motivation, I'll give these ideas
   a  more  concrete form, by analyzing a particular connec-
   tionist cognitive map (incl.  its  construction,  mainte-
   nance, and use) in pre-conceptual terms.



Dynamic Rate Adaptation.
Dr. Garrison W. Cottrell Department of Computer Science  and
Engineering University of California, San Diego, USA

   We describe a technique for automatically  detecting  the
   rate of an incoming signal. We first build a model of the
   signal using a recurrent network trained to  predict  the
   input  at some delay, for a "typical" rate of the signal.
   Then, fixing the weights of this network,  we  adapt  the
   time  constant of the network using the derivative of the
   error with respect to the  time  constant,  adapting  the
   delay appropriately as well. We have found that on simple
   signals, the network adapts rapidly to new inputs varying
   in  rate from being twice as fast as the original signal,
   down to ten times as slow. We discuss  the  possibilities
   of the application of this idea to speech.  This is joint
   work with Fu-Sheng Tsung and Mai Nguyen.


>From Theory to Practice: A Case Study
Dr. David E. Rumelhart, Professor Stanford University

   Work in neural networks can be classified (roughly)  into
   three  classes:  (1)  Theory development: the development
   and analysis of the conceptual tools of neural  networks.
   (2)  Applications: The application of a particular neural
   network system for solving a particular  problems.  These
   may  be particular problems in artificial intelligence or
   practical engineering problems.  (3) Cognitive/biological
   model  building  I  will  describe  a case study in which
   careful attention to  theoretical  issues  combined  with
   results  from  cognitive/biological models have been use-
   fully employed in the development of system for recogniz-
   ing continuous cursive script.


                        The Sessions


                        Wednesday 9th

Session 1:      Opening / Invited Papers (Room 1)
Chair:  Lars Niklasson (SNCC-92 organiser)
08.30   Opening
09.00   Connectionism and the Mind-Body Problem:
        Exposing the Rift between Mind and Cognition
        Dr. Tim van Gelder, Assistant Professor,
        Dept. of Philosophy, Indiana University, USA
09.50   Explaining Cognition with Nonlinear Dynamics
        Dr. Jordan B. Pollack, Assistant Professor,
        Laboratory for AI Research, Ohio State Univ, USA
10.40   Coffee Break



Session 2:      Invited Paper (Room 1)
Chair:  Dr. Tim van Gelder (Indiana University, USA)
11.10 - 12.00 Semantic and Syntactic Decompositions of
        Fully Distributed Representations
        Dr. Noel E. Sharkey, University of Exeter, UK



Session 3a:     Philosophical presentations (Room 1)
Chair:  Dr. Tim van Gelder (Indiana University, USA)
12.05   Subsymbolic Connectionism: Representational Vehicles
        and Contents
        Tere Vaden, Dept. of Mathematical
        Sciences/Philospohy, University of Tampere, Finland
12.30   First Connectionist Model of Nonmonotonic Reasoning:
        Handling Exceptions in Inheritance Hierarchies
        Mikael Boden, Connectionist Research Group,
        University of Skovde and Dr. Ajit Narayanan, Dept.
        of Computer Science, University of Exeter, UK
12.55   Lunch Break



Session 3b:     Theoretical presentations (Room 2)
Chair:  Dr. Jordan B. Pollack (Ohio State University, USA)
12.05   Neural Networks for Unsupervised Linear
        Feature Extraction
        Reiner Lenz and Mats Osterberg Image Processing
        Laboratory, Dept. EE. Linkoping University
12.30   Feed-forward Neural Networks in Limiting Cases of
        Infinite Nodes
        Abhay Bulsari and Henrik Saxen, Abo Akademi, Finland
12.55   Lunch Break



Session 4:      Invited Papers (Room 1)
Chair:  Dr. Jerome A. Feldman (ICSI, Berkeley, USA)
14.00   SCRuFFy: An Applications-oriented Hybrid
        Connectionist/Symbolic Shell
        Dr. James A. Hendler, Associate Professor,
        Dept. of Computer Science, UMCP, USA
14.50   Neurocomputing Hardware: Present and Future
        Dr. Dan Hammerstrom, Founder and Chief Technical
        Officer Adaptive Solutions, Inc., USA
15.40   Coffee Break



Session 5a:     Philosophical presentations (Room 1)
Chair:  Dr. Noel E. Sharkey (University of Exeter, UK)
16.10   Connectionism - The Miracle Mind Model
        Lars Niklasson, Connectionist Research Group,
        University of Skovde and Dr. Noel E. Sharkey, Centre
        for Connection Science, University of Exeter, UK
16.35   Some Properties of Neural Representations
        Christian Balkenius, Cognitive Science,
        Dept. of Philosophy, Lund University
17.00   Behaviors, Motivations, and Perceptions In
        Artificial Creatures
        Per Hammarlund and Anders Lansner, SANS, NADA,
        Royal Institute of Technology, Stockholm
17.25   Break




Session 5b:     Hardware-oriented presentations (Room 2)
Chair:  Dr. Dan Hammerstrom (Adaptive Solutions Inc., USA)
16.10   Pulse Coded Neural Networks for Hardware
        Implementation
        Lars Asplund, Olle Gallmo and Ernst Nordstrom, Dept.
        of Computer Systems, Uppsala University and
        Mats Gustafsson, Dept. of Technology, Circuits and
        Systems Group,  Uppsala University
16.35   Towards Modular, Massively Parallel Neural Computers
        Bertil Svensson, Dept. of Computer Engineering,
        Chalmers University of Technology, Goteborg and
        Centre for Computer Science, Halmstad University,
        Tomas Nordstrom, Div. of Computer Science and
        Engineering, Lulea University of Technology,
        Kenneth Nilsson and Per-Arne Wiberg, Centre for
        Computer Science, Halmstad University
17.00   The Grid - An Experiment in Neurocomputer
        Architecture
        Olle Gallmo and Lars Asplund, Dept. of Computer
        Systems, Uppsala University
17.25   Break


Session 6a:     Theoretical presentations (Room 1)
Chair:  Dr. Garrison W. Cottrell (University of California,
USA)
17.40   A Neural System as a Model for Image Reconstruction
        Mats Bengtsson, Swedish Defence Research
        Establishment, Linkoping
18.05   Internal Representation Models in Feedforward
        Artificial Neural Networks
        Hans G. C. Traven, SANS, NADA, Royal Institute of
        Technology, Stockholm
18.30 - 18.55   A Connectionist Model for Fuzzy Logic,
        Abhay Bulsari and Henrik Saxen, Abo Akademi, Finland



Session 6b:     Application-oriented presentations (Room 2)
Chair:  Dr. James A. Hendler (University of Maryland, USA)
17.40   A Robust Query-Reply System Based on a Bayesian
        Neural Network
        Anders Holst and Anders Lansner, SANS, NADA,
        Royal Institute of Technology, Stockholm
18.05   Neural Networks for Admission Control in an
        ATM Network
        Ernst Nordstrom, Olle Gallmo, Lars Asplund,
        Dept. of Computer Systems, Uppsala University
18.25 - 18.55   A connectionist model of learning novel
        words
        Amanda Sharkey, Centre for Connection Science,
        University of Exeter, UK



Swedish Neural Network Society (SNNS) (Room 1)
19.00-19.45     Members meeting



                Thursday 10th

Session 7:      Invited Papers (Room 1)
Chair:  Dr. Ronald L. Chrisley (University of Sussex, UK)
09.00   Structure and Change in Connectionist Models
        Dr. Jerome A. Feldman, ICSI, Berkeley, USA
09.50   On Nativist Connectionism
        Dr. Ajit Narayanan, Dept. of Computer Science,
        University of Exeter, UK
10.40   Coffee Break



Session: 8      Invited Paper (Room 1)
Chair:  Dr. Anders Lansner (Royal Institute of Technology,
Stockholm)
11.10 - 12.00   The Neuropharmacology of Associative Memory
        Function: an in Vitro, in Vivo, and in Computo
        Study of Object Recognition in Olfactory Cortex
        Dr. James M. Bower, Associate Professor, Div. of
        Biology, California Institute of Technology, USA




Session 9a:     Neurobiological presentations (Room 1)
Chair:  Dr. James M. Bower (CalTech)
12.05   A Model of Cortical Associative Memory Based on
        Hebbian Cell Assemblies
        Erik Fransen, Anders Lansner and Hans Liljenstrom,
        SANS, NADA, Royal Institute of Technology, Stockholm
12.30   Cognition, Neurodynamics and Computer Models
        Hans Liljenstrom, SANS, NADA, Royal Institute of
        Technology, Stockholm
12.55   Lunch Break



Session 9b:     Application-oriented presentations (Room 2)
Chair:  Dr. David E. Rumelhart (Stanford University)
12.05   Experiments with Artificial Neural Networks for
        Phoneme and Word Recognition
        Kjell Elenius, Dept. of Speech Communication
        and Music Acoustics, Royal Institute of
        Technology, Stockholm
12.30   Recognition of Isolated Spoken Swedish Words -
        An Approach Based on a Self-organizing Feature Map
        Tomas Rahkkonen, Telia Research AB, Systems Research
        Spoken Language Processing, Haninge
12.55   Lunch Break



Session 10:     Invited Papers (Room 1)
Chair:  Dr. Ajit Narayanan (University of Exeter)
14.00   Connectionist Cognitive Maps and the
        Development of Objectivity
        Dr. Ronald L. Chrisley, School of Cognitive and
        Computing Sciences, University of Sussex, UK
14.50   Dynamic Rate Adaption
        Dr. Garrison W. Cottrell, Dept. of Computer
        Science and Engineering, University of California,
        San Diego, USA
15.40   Coffee Break



Session 11:     Invited Paper (Room 1)
Chair:  Mikael Boden (SNCC-92 organiser)
16.10   From Theory to Practice: A Case Study
        Dr. David E. Rumelhart, Professor,
        Stanford University, USA
17.00   Closing




                     Registration form

Fees include admission to all conference sessions and a copy
of the Advance Proceedings.

Hotel reservation is made at Billingehus Hotel  and  Confer-
ence  Centre.  The  rooms  are  available from Tuesday (8th)
evening to noon Thursday (10th), for  the  two-day  alterna-
tive,  and  from  Wednesday  evening  (9th) to noon Thursday
(10th), for the one-day alternative. All activities will  be
held  at  the  conference  centre. A block of rooms has been
reserved until 10th Aug.  After this date, room reservations  
will be accepted  on  a  space available basis. To register, 
complete and return the form below to the secretariat.

Registration is valid  when  payment  is  received.  Payment
should  be  made  to  postal  giro  78 81 40 - 2, payable to
SNCC-92, Hogskolan i Skovde. Cancellation of hotel  reserva-
tion can be made until 15/8 (the conference fee, 600 SEK, is
not refundable).


-------------------------- cut -----------------------------
Name: (Mr/Ms)   ____________________________________________
Company:        ____________________________________________
Address:        ____________________________________________
City/Country:   ____________________________________________
Phone:          ____________________________________________
If the double room alternative has been chosen, please give
the details for the second person.
Name: (Mr/Ms)   ____________________________________________
Company:        ____________________________________________
Country:        ____________________________________________

Alternatives                       Until 9th Aug After 9th Aug
(please circle chosen fee)

Conference fee only                1000SEK       1000SEK
(incl. coffee and lunch)

Conference fee + Full board and
single room lodging 8/9 - 10/9     2400SEK       3500SEK

Conference fee + Full board and
double room lodging 8/9 - 10/9     2 x 2000SEK   2 x 2600SEK

Conference fee + Full board and
single room lodging 9/9 - 10/9     1700SEK       2400SEK

Conference fee + Full board and
double room lodging 9/9 - 10/9     2 x 1500SEK   2 x 2000SEK


Indicate if vegetarian meals are preferred: _____ person(s)


