~/writing/math-foundations-for-reasoning-models $ ls
Computation as State Transition
Move Operation
A move operation changes the current tape position without changing the tape contents.
~/writing/math-foundations-for-reasoning-models $ cat topic.md
What Is a Move Operation?
Definition
A move operation changes the current tape position without changing the tape contents.
Move does not read a symbol.
Move does not write a symbol.
Move only changes where the system is focused.
Move = change the current tape position
Move vs Tape
A tape stores symbols in ordered positions.
A move operation changes which position is current.
Example:
Tape:
[1] [0] [1]
^
current position
Move right:
[1] [0] [1]
^
current position
The tape contents stayed the same.
Only the current position changed.
Tape stores symbols. Move changes where the system is focused on the tape.
Basic Example
Tape:
[1] [0] [1]
^
current position
Move right:
[1] [0] [1]
^
current position
Move left:
[1] [0] [1]
^
current position
The symbols did not change.
The active position changed.
Move changes position, not tape contents.
Move vs Read And Write
Read, write, and move do different things.
Read -> observe the current symbol
Write -> replace the current symbol
Move -> change the current position
Example:
Tape:
[1] [0] [1]
^
Read result:
0
Write:
1
Tape after write:
[1] [1] [1]
^
Move right:
[1] [1] [1]
^
Read made the current symbol available.
Write changed the current symbol.
Move changed where the next operation will happen.
Move does not inspect or modify symbols. It changes where later operations apply.
Current Position Is Part Of State
The current position is part of the system's state.
Example:
Tape contents:
[1] [0] [1]
First position state:
current position = first cell
Second position state:
current position = second cell
The tape contents are the same.
But the system state is different because the current position is different.
This matters because the next read or write will affect a different cell.
Moving changes the system state by changing the current position.
Move Changes Future Reads And Writes
Move does not change symbols directly.
But it changes which symbol future operations will access.
Example:
Initial:
[1] [0] [1]
^
Move right:
[1] [0] [1]
^
Next read:
1
Without the move, the next read would have returned 0.
After the move, the next read returns 1.
Move affects computation by changing where future reads and writes happen.
Move Direction
In a simple tape model, movement usually has directions.
Common directions:
- left
- right
- stay in place
Example:
Move left:
[1] [0] [1]
^
Move right:
[1] [0] [1]
^
Some models allow only left or right.
Some also allow staying at the same position.
Move direction defines how the current position changes.
Boundary And Extendable Tape
Some tapes are treated as extendable.
Example:
[1] [0] [1] [_] [_] [_] ...
Moving right can reach unused blank positions over time.
But this does not mean every model has the same boundaries.
Some systems may have fixed memory limits.
Some theoretical models treat the tape as extendable without a fixed limit.
Movement depends on how the tape model defines its boundaries.
Move Is Chosen By A Rule
Move does not decide where to go by itself.
A transition rule chooses the move action.
Example:
If read symbol is 0:
write 1
move right
If read symbol is 1:
write 0
move left
Suppose the read symbol is:
0
Applicable move action:
move right
The move operation changes the current position.
The rule determines which movement is valid.
The rule chooses the movement. Move updates the current position.
Key Properties
- Changes the current tape position
- Does not change tape contents
- Does not create, remove, or replace symbols
- Changes the system state because current position is part of state
- Affects where future reads and writes happen
- Is selected by a transition rule
Move changes where the system operates, not the stored symbols.
Role in Computation
The move operation allows the system to:
- reach different tape positions
- read from new positions
- write to new positions
- process ordered memory step by step
Accessing Memory
By moving step by step, the system can reach other tape positions over time.
Example:
[1] [0] [1] [_]
^
[1] [0] [1] [_]
^
[1] [0] [1] [_]
^
The system interacts with one position at a time.
Movement allows it to process a larger tape gradually.
Local vs Global Access
Access is local.
The system is focused on one position at a time.
But repeated movement lets the system work across many positions over time.
A large tape can be processed through repeated local movement.
Why This Matters
Without move:
- the system is limited to one tape position
- it cannot scan across stored symbols
- it cannot process ordered memory beyond the current cell
With move:
- the system can scan stored symbols
- the system can update different cells
- the system can build or inspect structures across the tape
Move gives local operations reach across ordered memory.
Final Summary
- Move changes the current tape position.
- Move does not change tape contents.
- Move does not create, remove, or replace symbols.
- Current position is part of the system state.
- Move changes where future reads and writes happen.
- Move direction defines how the position changes.
- Movement depends on the tape model's boundaries.
- The transition rule chooses the move action.
- A large tape can be processed through repeated local movement.
~/writing/math-foundations-for-reasoning-models $ open
Continue the series.
Finite Control