Mystery Number Five - Computational Linguistics and Talking Robots - page 91

The sequences are performed by the following LA-grammar, called LA-act1
(for action). The definition begins with the variables:

6.1.5 V ARIABLE DEFINITION OF THE LA- ACT 1 GRAMMAR
T n {red, green, blue} and n {1, 2, 3, ...}
M1 {strght, left, right}
M2 {strght, left, right}
K

{x i ,y i ,z i , ...} and i {1, 2, 3, ...}

The variables are restricted to the values in the corresponding sets. The rule
system of LA-act1 is defined as follows:

6.1.6 R ULE SYSTEM OF THE LA- ACT 1 GRAMMAR
ST S = def

{( rec: T n {Rule_0, Rule_1}) }

Rule_0 {Rule_0, Rule_1}

⎡ ⎤ ⎡ ⎤

rec: T n
next: T n +1
prn: K i

rec: T n +1
prev: T n
prn: K i +1

⎣ ⎦ ⎣ ⎦

output position nw

Rule_1 {Rule_2}

⎡ ⎤ ⎡ ⎤

rec: T n
next: M1
prn: K i

act: M1
prev: T n
prn: K i +1

⎣ ⎦ ⎣ ⎦

output position nw

Rule_2 {Rule_2}

⎡
⎣

⎤
⎦

⎡
⎣

⎤
⎦

act: M1
next: M2
prn: K i

act: M2
prev: M1
prn: K i +1

output position nw

{( next: rp Rule _ 2 )} ST F = def

The LA-act1 grammar automatically starts navigating through the agent's
Word Bank whenever its start state ST S is satisfied. This is the case when-
ever the agent's recognition provides one of the values of the trigger variable
T n (6.1.5), i.e., whenever the agent perceives a red, green ,or blue light.
The rule package of the start state ST S calls Rule_0 and Rule_1. Rule_0 may
be applied recursively to parse complex stimuli, like red green or red red
red . Rule_1 moves from the end of the stimulus to the beginning of the mo-
tion sequence and calls Rule_2. Rule_2 calls itself and completes the motion
sequence. After that the fixed behavior agent comes to rest until it is triggered
by another stimulus.
As an example of a rule application consider Rule_1, called from the start
state (rather than from Rule_0) and applied to the input proplet red :