Appendix: Word Boundaries

The Multi-Stroke Dilemma

If Plover only supported single-stroke words, then it would be obvious where each word ended and the next began. But with multi-stroke words, Plover has to decide whether a given stroke is a whole word or just a part of a larger word. It needs to make this decision partly because it wants to output the correct word, and partly because Plover inserts spaces automatically between words, and needs to get that right as well.

Plover (and all steno systems) must make this choice for each stroke you type. This introduces a whole new class of homophones into the mix. If Plover only supported single stroke words, it would only need to disambiguate between individual words like 'beat' and 'beet'. By supporting multi-stroke words, Plover also has to deal with homophonic word sequences like 'light engage' and 'lighten gauge'.

The Natural Resolution

Plover can't intuitively know which homophonic word sequence you intend with a series of strokes. But it does have a consistent and predictable set of internal rules that results in a clean selection for every sequence of strokes:

  1. Plover will never split a word it has already output. For example, once Plover has output the word 'lighten', a later stroke can never cause Plover to back up and split that word into 'light' and another word beginning with 'en-'.
  2. Plover will always engulf a whole word it has already output into a longer word, if available. For example, if you stroke TPAOEUR, Plover will immediately output the word 'fire'. If you then stroke PHRAEUS, Plover will remove the word 'fire', and engulf that stroke into the new word 'fireplace' instead.
  3. Plover will always choose the longest available series of strokes that match a dictionary entry. This is really just another way to say (2) above. For example, if you type a series of 10 strokes, you might see Plover output 10 different words. But if, on the 11th stroke, Plover sees a matching 11-stroke entry in your dictionary file, it will backspace over all 10 of those previous words, and replace them with the word that matches that 11-stroke entry.

Taking Control

Plover supports various techniques for controlling which multi-stroke homophone it should output. The most preferred option by far is described in Lesson 5: Prefix And Suffix Strokes. But there are other techniques that, while perhaps more convenient for the beginner, are slow and should not be considered a real solution to the problem.