But can copies be their own different, separate and independent entities?
Yes. Your example of cell division demonstrates this quite nicely.
A meditation on filesystems
Imagine copy-pasting a folder named A. We copy and paste A and name the copy B. Is B A and is A B?
It depends what you mean by “copy”. There are at least six different kinds of copy that you can have in a computer filesystem:
- Hard link
- Symbolic link
- Shallow copy
- Overlay mount
- Deep copy
- Copy-on-Write
Hard link
A directory looks a bit like this:
- This is a directory.
.
is stored at 0x762920
..
is stored at 0x762736
A
is stored at 0x240008
readme.txt
is stored at 0x763424
The hexadecimal codes, like 0x240008
, refer to the actual location on the hard drive that things are stored. The directory itself is just data stored at some location of the drive; in this example, the parent is stored at 0x762920
.
If you made a hard link to A
called B
, the parent directory would then look like this:
- This is a directory.
.
is stored at 0x762920
..
is stored at 0x762736
A
is stored at 0x240008
readme.txt
is stored at 0x763424
B
is stored at 0x240008
Here, A
and B
are essentially two different labels for the same directory. If you deleted A
, the directory would still be accessible via B
. Two folder icons would be visible in your file manager. Modifying a file in B
would modify the same file in A
; creating a file in A
would create the same file in B
.
Symbolic link
Consider the directory from before. Making a symbolic (soft) link to A
called B
would look like this:
.
is stored at 0x762920
..
is stored at 0x762736
A
is stored at 0x240008
readme.txt
is stored at 0x763424
B
can be found at A
If you deleted B
, the directory would still be found at A
– but if you deleted A
, B
would point nowhere. Two folder icons would be visible in your file manager, though B
might have a little “link” icon on it. Modifying a file in B
would modify the same file in A
; creating a file in A
would create the same file in B
.
Shallow copy
In a shallow copy, the parent directory would look like this:
- This is a directory
.
is stored at 0x762920
..
is stored at 0x762736
A
is stored at 0x240008
readme.txt
is stored at 0x763424
B
is stored at 0xA36092
but the directory entries stored at 0xA36092
would be the same as the ones stored at 0x240008
– at least initially. In effect, a new directory has been created, and a hard link at 0xA36092
has been created for every file in A
. Modifying a file in B
would modify the same file in A
, but creating a file in A
would not create the same file in B
.
Overlay mount
A new, empty directory B
would be created:
- This is a directory
.
is stored at 0x762920
..
is stored at 0x762736
A
is stored at 0x240008
readme.txt
is stored at 0x763424
B
is stored at 0xA36092
Then the overlay mount B
-over-A
would be created. Two folder icons would be visible in your file manager, though B
might have a little mount icon on it. Modifying a file in A
would modify the corresponding file in B
, creating a file in A
would create the corresponding file in B
, but creating a file in B
would not create the corresponding file in A
.
Deep copy
In a shallow copy, the parent directory would look like this:
- This is a directory
.
is stored at 0x762920
..
is stored at 0x762736
A
is stored at 0x240008
readme.txt
is stored at 0x763424
B
is stored at 0xA36092
However, every file and folder inside B
would be a deep copy of the corresponding file and folder in A
, so they would also be stored in a separate location. Modifying or creating files in A
would not affect what happens in B
, or vice versa.
Copy-on-Write
Here, the directory entry looks a little different:
- This is a directory with 1 reference.
.
is stored at 0x762920
..
is stored at 0x762736
A
is stored at 0x240008
readme.txt
is stored at 0x763424
When you copy A
to B
, it creates a hard link:
- This is a directory with 1 reference.
.
is stored at 0x762920
..
is stored at 0x762736
A
is stored at 0x240008
readme.txt
is stored at 0x763424
B
is stored at 0x240008
When it creates this hard link, it also increases the number of references to A
and B
(and to everything inside them) by one. When you modify anything with more than one reference, it subtracts one from the reference, creates a copy, and then moves the directory entry in the thing's parent directory to point to that copy. (If the parent directory has more than one reference, then it gets copied, and its directory entry in its parent gets modified to point to the copy.) So if you subsequently created a file in A
, the parent would then look like:
- This is a directory with 1 reference.
.
is stored at 0x762920
..
is stored at 0x762736
A
is stored at 0xA36092
readme.txt
is stored at 0x763424
B
is stored at 0x240008
However, if a cosmic ray flipped some bits on the hard drive before either A
or B
was modified, both of them would change.
What does this mean for philosophy?
You have an imprecise concept in your head, and are confused that multiple interpretations of that vague idea conflict with each other.
What if we copy-paste the A folder and name the copy A? That doesn't mean A is in two places at once, right? That if I delete one of the As, the deleted A still exists?
It depends. But if you pick a specific meaning of “copy”, the answer is easy to discover.
In the real world, things are a lot easier. When you make an actual copy of something, you end up with a separate object; the original and the copy aren't magically linked to each other, so you can dip one in paint without affecting the other.
- They are the same, until they're not (but due to real-world physics, even a magical duplicate won't stay identical for more than the briefest moment).
- They are separate, because they are not connected.
- They are independent, because the state of one does not depend on the state of the other.
This is not necessarily true for data stored in computer file systems (or for parts of quantum systems, but for different reasons); nonetheless, these three things are true for normal, classical facsimiles.