Figuratively Speaking - Natural and Artificial Reasoning - page 173

Actor 2
1[ 0.060 0.113 0.113 0.178 0.310 0.113 0.113 ] I 0.0.161
2[ 0.132 0.128 0.207 0.192 0.114 0.128 0.099 ] I ] 0.148
Actor 3
1[ 0.060 0.113 0.113 0.113 0.376 0.113 0.113 ] I 0.0.171
2[ 0.132 0.128 0.128 0.276 0.109 0.128 0.099 ] I ] 0.152
(same 2)
Actor 1
1[ 0.091 0.164 0.122 0.239 0.127 0.127 0.129 ] I 0.0.149
2[ 0.211 0.123 0.197 0.114 0.128 0.128 0.099 ] I ] 0.148
Actor 2
1[ 0.097 0.127 0.121 0.159 0.239 0.127 0.129 ] I 0.0.149
2[ 0.132 0.128 0.207 0.192 0.114 0.128 0.099 ] I 0.0.148
Actor 3
1[ 0.097 0.127 0.127 0.116 0.276 0.127 0.129 ] I 0.0.152
2[ 0.132 0.128 0.128 0.276 0.109 0.128 0.099 ] I ] 0.152
In the final modification after seven more cycles we have:
Actor 1
1[ 0.089 0.167 0.131 0.198 0.136 0.138 0.142 ] I 0.0.146
2[ 0.180 0.132 0.191 0.126 0.138 0.139 0.094 ] I ] 0.146
Actor 2
1[ 0.093 0.139 0.133 0.160 0.198 0.136 0.142 ] I

0.146
2[ 0.145 0.137 0.172 0.186 0.126 0.139 0.094 ] I ] 0.146
Actor 3
1[ 0.093 0.138 0.138 0.127 0.227 0.135 0.142 ] I 0.0.148
2[ 0.145 0.138 0.135 0.227 0.121 0.139 0.094 ] I ] 0.148
Given the limited experience of each Actor this shows that there is the possibil-
ity of convergent internal standards. The limitation is that the upper bounds of the
evaluation of the objects cannot be shifted until new and 'better' objects have been
experienced. Actor 1 has inferred correctly the relative position of object 2 with re-
spect to object 1 and Actor 2 has adopted an evaluation of object 2 that is approaching
a consensus.

=

13.11 Future Development

This initial experiment with this model was only intended to establish the possibility
of inferred internal referencing. Experiments that are more extensive will need to be
done to show what might be the limitations and stability of such internal and inferred
references. More Actors experiencing a wider range of objects over an extended set
of events is required. Investigations into the way in which the internal referencing
might change are needed as well as the seeing if there is the possibility of distinct
sub-groups of Actors appearing due to drift. Such drift would be expected as soon
as the Actors become independent of the forced consultation sequence used here.
The model could be extended along the lines of Gooding and Addis ( 1999 —also
see Chap. 7). We would intend to use the development of this model to revisit our
observations and we would expect this to lead to the construction of new experiments.