Get a list of all dictionary words in English:

Get a list of words that match the pattern of starting with “z” and then having 3 more letters:

This finds 5 words nearest to “elephant”, based on changing the fewest letters:

This finds all the words in the dictionary that differ from “fish” by at most 1 letter:

Use Rest to keep the “rest” of the list, dropping the original word:

We’re going to want to create networks of words. Let’s start by creating a network of numbers.

This defines a particular way to connect numbers from 0 to 9:

Graph draws a network or “graph”; VertexLabels->“Name” says to label vertices with their names:

This creates a list of words that are going to be in the network, here consisting of “q” followed by 3 letters:

We’re naming this list words.

Now we construct a function that finds nearest words:

We can apply this function to any word in the list, saying we want all words that differ by at most 1 letter:

Use Rest to keep only the “rest” of the list, dropping the first element (which is here the original word):

To find the whole network, we need to apply this to every word in the list:

The # and & set up a pure function, which is then mapped over the original list of words.

To form the actual network, we need to “thread” these connections. Thread turns a connection to a list into a list of connections—for example:

Thread the nearest connections:

Then flatten out all the sublists:

Now we can turn this into a graph, or network:

This creates a network (named g) of nearest words for all 4-letter dictionary words (note that there’s a semicolon at the end to avoid printing the very big result):

This finds the shortest path on the network from “fish” to “fowl”—a word ladder: