A collation is a structured alignment with certain placeholders that underpins the construction of formal languages.
These placeholders may be replaced by elements of an alphabet $\mathcal A$ under consideration.
A collation in $\mathcal A$ is one where all placeholders are replaced by symbols from $\mathcal A$.
For example, if we take $\square$ to denote a placeholder, then $\square\square\square\square\square$ represents the collation "a word of length $5$".
We can see that then the word "sheep" is an instance of the collation "a word of length $5$" in the English alphabet, as is "axiom".
A key feature of collations is the presence of methods to collate a number of collations into a new one.
A collection of collations, together with a collection of such collation methods may be called a collation system.
- If $C$ and $C'$ are indistinguishable, then $C = C'$.
Then $\mathcal C$ has the unique readability property for $\mathcal A$.
The concept of collation being a very fundamental and abstract one, it is helpful to discuss some examples.
- Any word in natural language is a collation in the standard alphabet;
- Any number is a collation in the alphabet of digits;
- Any sentence is a collation in the alphabet of all words;
- Any sentence is a collation in the alphabet of letters and punctuation marks;
- Any movie is a collation in the alphabet of stills;
- Any Lego construction is a collation in the alphabet of Lego bricks;
- Any labeled tree is a collation.
We see that unique readability is typically ensured by:
- Position on the paper (or any other 2D carrier);
- Position in time;
- Position in the real world (or any other 3D environment).