-
Notifications
You must be signed in to change notification settings - Fork 277
Glossary
This glossary will help you with terms that you may come across in the steno community.
Table of Contents
- Brief
- Chord
- Conflict
- Dragging
- JSON
- RTF/CRE
- Misstroke
- Stacking
- Steno Dictionary
- Steno Order
- Stroke
- Steno Theory
- Untranslate
- Word Boundary
Also known as the "abbreviation", "short form", or "arbitrary". Briefs are simply non-phonetic mappings of steno outlines to English words or phrases. Common words and phrases are often briefed for the purpose of speed. For instance, the phrase "from time to time" would regularly be written out:
FROM/TAOIM/TO/TAOIM
(reads: "from/time/to/time" and takes four strokes)
Or, as a simple brief:
FRIMT
(reads: "frimt" and takes only one stroke)
The pressing down of multiple keys at the same time. Contrast with a QWERTY-style typing system which hits only one key at a time.
-
traditionally: a conflict-based theory uses one stroke for multiple translations. For example, a non-realtime stenographer could use the same stroke for "bare", "bear", and "bar", which is a conflict that the stenographer would have to manually resolve at a later time. Plover is a realtime-only system and does not support conflict-theories.
-
informal; a.k.a collision: when two dictionaries have the same stroke, the dictionary with the highest priority is favored. For this reason, it is important to understand your dictionary order.
Dragging is the term used to describe accidentally dragging another key into your stroke. E.g. if you try to write -F
but then drag your finger to the left, you might hit *F
instead.
JSON, in the context of stenography, is a dictionary format which maps steno
strokes to translations. You will often see strokes expressed in the JSON format,
such as "SKP": "and"
.
A misstroke is like a "chord typo". It's when you mean to write one chord, but stroke another. Often, dictionaries have misstroke entries that are added when a stenographer frequently misstrokes an entry. For example, take the stroke TKPWAOD
(meaning GAOD
) which translates to good
. Sometimes the stenographer may miss a key, so they could have a misstroke entry TKPAOD
which would also translate to good
. Then they are protected from these typos in regular writing. There are many misstroke entries in the default dictionary, and you must try to make sense of results when you look up words, instead of blindly accepting the shortest stroke.
Plover can read steno dictionaries in JSON and RTF/CRE format. RTF/CRE stands for
rich text format with court reporting extensions
. It is a standard format that most
proprietary steno software can import from and export to. Plover can read RTF/CRE
natively.
When writing stenography, if you accidentally merge two strokes into one, it is called stacking. For example, you might try to write the rap
with T-/RAP
but end up with TRAP → trap
because you stacked the strokes. To avoid stacking, the stenographer must be sure to release all keys in their chord before stroking the next. Sometimes machines are prone to stacking due to bad debouncing or sticking keys.
Used by Plover or other stenotype software. Contains all the words and the strokes that produce those words. While generally these are constructed using a steno theory, this can be freely modified by the stenographer. Dictionaries are a collection of entries, which map strokes to translation.
The 22 keys on the steno machine has an explicit "order" that gets read out,
top-to-bottom, left to write. The entire steno layout is defined by STKPWHRAO*EUFRPBLGTSDZ
.
Can refer to a chord, or a set of chords that you have for a
given translation. E.g., "my stroke for steno-dude
is STOEUPB/TKAOUD
"
A "system" or way of thinking that determines which steno strokes will match to which words. Theories range generally from being based on spelling ("brief-heavy") to being based on the sound of the word ("stroke-heavy"). The dictionary included with Plover uses a theory based on is based on NYCI theory which is descended from StenEd. It offers a hybrid between a brief-heavy and stroke-heavy theory. It is recommended to start learning with Plover theory, and you will likely learn what style you like and you can always switch later. Mirabai uses the Plover dictionary professionally.
When writing in stenography, your strokes map to translations. E.g. KAT → cat
. However, if a stroke is not in your dictionary, the raw form will be outputted instead. This is called an untranslate. For example, if your dictionary doesn't have KAT
, Plover will simply output KAT
.
The implicit spacing in between words. Spacing is inserted automatically by Plover or other steno software. As words and phrases will often sound similar to others, a stenographer needs to choose the stroke or brief appropriate for the situation with the correct word boundary.
An illustration of a word boundary error is given by the phrase "cat log". If a stenographer were to write "cat log" with Plover, by default, the system will write "catalog". This happens because "cat log" isn't a very common word-pair in English. The stenographer must explicitly write "cat (space) log". But, there are many more common cases that are handled and the stenographer must be explicit. See below for how some phonetics are differentiated:
Examples (Plover):
- "in here";
TPH
/HAOER
vs "insect";EUPB
/SEBGT
- "on top of";
OPB
/TOP
/-F
vs "onto";AUPN
/TO
- "it is a live (wire)";
T
/S
/AEU
/HREUF
vs "it is alive";T
/S
/A
/HRAOEUF