Brian,

It's as simple as this. Take none, some, most, all and even character as levels on a ladder, with none being the lowest level. Now, let's say that we have the dash (-) sign. By default, this symbol is on the most level. So, it stands on the third level of the ladder. If I set my NVDA to speak punctuations on that level and any level above, it will speak that symbol. If I set NVDA to none or some, it won't. The level symbols that are pronounced is preset by NVDA, but you can customize these levels, and other features of symbols, such as the name, and if the symbol should be affected by the synthesizer. In the previous example, the dash (-) symbol is set to most, but to never be affected by the synthesizer. I set the dash so that it is always affected, so a pause sometimes occurs when it is encountered.

So, none, some, most, all and character are NVDA created preset levels, with symbols placed on each step or category.