To understand indexing, one of the things you need to understand is the concept of "context".

 

From the Shorter Oxford English Dictionary:

 

Context, noun

 

3. The part or parts immediately preceding or following a passage or word as determining or helping to reveal its meaning; the surrounding structure as determining the behaviour of a grammatical item, speech, sound, etc.  (16th century).

 

>b figurative: Ambient conditions; a set of circumstances; relation to circumstances.  (19th century).

 

[Latin contextus from context- past participle stem of contexere weave together, formed as CON- + texere weave.]

 

As part of understanding "context", it is also important to understand something about the way we think.  Let's take for instance the word "valve". When this word is said (or written) alone (i.e. taken OUT of context), some people might think of a water faucet, others might think of the generic idea that it is something that regulates flow on some channel (e.g. air, water), and it is not clear at all what actual concept is meant.  What kind of valve?.  How big is it?  But if you give it context, then you can get very precise about what comes to mind when the reader (or listener) encounters it.  Example:

 

 

 

 

Note that I just conveyed the concepts of 5 very different kinds of valves, simply by supplying their context.  Notice that I did this using one, or at most a small number of words.  In the process, I hope I have also conveyed WELL the concept of "context".  Reason:  this needs to be well understood to make high-quality indexes for documents, large and small.

 

 

 

Note that one of the tools I used in these contexts is hierarchy.  If we think of hierarchy as a "tree structure", the closer one gets to the "leaves" in the tree, the narrower the concept becomes.  Conversely, the closer one gets to the "trunk" of the tree, the broader the concept is.  So we consider that the "narrower" concepts are "contained by" the broader concepts.  And in this structure, the relationships between them are shown by the broader concepts being listed first, and the narrower concepts being listed with the broader concepts, but their indentation level shows that they are contained by with the broader concept, and are themselves narrower in scope.

 

Hierarchy in context is one of the powerful tools that can make an index more useful to your readers, and used well, will enable your readers to find things they are looking for quickly.

 

 

 

More complex indexes use tree-structured hierarchy like the above to show hierarchical relationships between topics.  In addition to regular "stand-alone" index entries that contain the entire topic as well as page number, you will often see general nouns listed (without page numbers) that have indented subentries that are more specific.  Such structures can be 2 levels, and sometimes even 3 levels deep.

 

An example you might see in the user's guide for a word processor:

 

 

Here, "editor window" is the general term.  Note that it is an index entry of its own, but does not have a specific page number associated with it.  It is there to provide context for the subentries.  Note that the subentries are listed in alphabetical order, and that they provide the actual page references.

 

For the above example, we say that the "editor window" entry is at "level 1" or is a "level-1 index entry", and its subentries are at "level 2" or are "level-2 index entries".  If there is a third level, these will be subentries for level-2 entries, which will provide context for the level-3 entries.

 

It is technically possible to have more than 3 levels, but for practical purposes, they are almost never seen in professional indexes.