Re: The "mini-chart" form

[leeo]

The minichart form is proposed as an alternative to the standard
"compressed papertape" form for a steno outline. Compare:

compressed papertape form:
HAOER/-S/APB/KPAFPL/-L/OF/THEUS (<-- here is an example of this)

the "minichart" form:
...h@'er..: ....__'...s ....a_'.n.: .ex.a_'~mp: ....__'..l:
...._o'f..: .t.h_i'...s

Though it requires more characters, on the average, it serves three
purposes: (1) Being closer to what is actually stroked, it is easier
to envision the action it represents without being as cumbersome as a
complete stenographer's chart. This makes it useful as a study guide.
(2) it gives clients a "best guess" pronunciation for a stroke
sequence not found in the dictionary. "...h@e'r..: ....__'...s
....a_'.n.: .ex.a_'~mp: ....__'..l:" is easier to decipher for the
uninitiated than "HAOER/-S/APB/KPAFPL/-L". (3) it provides a
consitency check of the theory. The requirement is that any
substutions be strictly theory related; no dictionary is permitted.

The common "chords" that always represent definite sounds are
translated immediately. For example, the /TP-/ combination represents
the initial "f" sound, so a word like "fat", /TPAT/ appears as
.~f.a_'...t

The notation is positional, where each of the eleven characters in the
minichart represents a particular column of keys on the standard steno
keyboard. ....__'...: indicates an "untouched keyboard". The colon at
the far right emphasizes the multiple positions and the two columns
for the right little finger. The untouched vowel keys appear as "_".
The empty asterisk key appears " ' ", and is placed after the vowels,
because the asterisk key is commonly hit with the right index finger
using a "wide asterisk" key.

Each of the keys of course has its standard name

s...__'...:
.t..__'...: .k..__'...:
..p.__'...: ..w.__'...:
...h__'...: ...r__'...:
....a_'...: ....o_'...:
....__*...:
...._e'...: ...._u'...:
....__'f..: ....__'r..:
....__'.p.: ....__'.b.:
....__'..l: ....__'..g:
....__'...t ....__'...s
....__'...d ....__'...z

Since each finger can also hit the "crack" between the two rows of
keys,
that is given the following designation

.d..__'...: [TK]
..b.__'...: [PW]
...l__'...: [HR]
....@_'...: [AO] -- U+0040 At sign
...._i'...: [EU]
....__'æ..: [FR] -- U+00E6 Latin small letter AE
....__'.n.: [PB]
....__'..ç: [LG] -- U+00E7 Latin Small Letter C with cedilla


The right little finger has four buttons and four combination options:
....__'...¢ [TS] -- U+00A2 Cent sign
....__'...$ [SZ] -- U+0024 Dollar sign
....__'...þ [DZ] -- U+00FE Latin small letter Thorn
....__'...đ [TD] -- U+0111 Latin small letter D with stroke

There are theoretically seven other "combinations" of the four keys
lying under the little finger that usually are not included in a
theory, because it would require that a hand leaves the "home
position".

*

Now, where this really shines is when "chord" represents a
particular sound. for instance for the soft-"G" sound, I don't think
/TKPW-/ or".db.__'...:", but ".~g.__'...:". In about 50 lines of
code, I make the letter substitutions for most of the theory elements,
such as ".db.__'...:" to ".~g.__'...:".

The chart doesn't specify which key row is hit, just the result from
its column of keys: Here is the alphabet as it appears with leading
and following forms where available:

....a_'...: the long a sound is ....ai'...:
..b.__'.b.:
.k.r__'...: -- c is infrequent enough not to substitute.
.d..__'...d
...._e'...: the long e sound is ....@e'...:
.~f.__'f..: --initial "f" is made with both upper keys
.~g.__'..g: --initial initial "g" made with both "cracks"
...h__'...:
...._i'...: the long i sound is ....@i'...:
~~~j__'.~j:
.k..__'.~k:
...l__'..l:
..~m__'.~m:
.~~n__'.n.:
....o_'...: the long o sound is ....oe'...:
..p.__'.p.:
.kw.__'...: -- I find keeping kw is more readable than "~q"
...r__'r..:
s...__'...s
.t..__'...t
...._u'...: the long u sound is ....@u'...:
~..v__'~v.: -- final -v is accomplished with "-FB" in my theory
..w.__'...:
.kp.__'.~~x -- with an included vowel, displays like ".ex._a'.~m.:"
.~~y__'...:
~.~z__'...z -- initial -z- accomplished with /SWR/ in my theory

in addition to the alphabet, common blends and chords are also
translated, provided they apply universally in the theory: For an
example, the final sound combination -sk is stroked /-FBG/, so any
combination with /-FBG/ appears in the mini-chart as ....__'s~k:, such
as /WEUFBG/ for ..w._i's~k: . If a theory element is not universally
applied, it is usually left untranslated. Consider the final
combination -GT often represents the "-th" sound, but there are enough
exceptions, such as /AIGT/ for "eight", "....ai'..gt", that this is
always left as "gt", like ..b.a_'..gt for "bath". If a theory element
would be too awkward to indicate, it is also left untranslated, like
where an asterisk key is incorporated, such as ..b._e*...s for "best",
and ..bla_*.ng: for "blank". Also, translations are deliberately
suppressed in the case of punctuations or other special functions.
A common stroke for the "." (period) punctuation is /PH-FP/, which
would normally translate to "..~m__'ch..:", but on its detection, it
is suppressed to simply read "..ph__'fp.:".

I make use of the unicode characters to make the chart readable and
fit in the outline "footprint", such as /-FR/ appearing as ....__'æ..:

All this rigamarole does have an important use: The requirement for
many real-time systems is to make a "best guess" for a stroke sequence
not in the dictionary, hence many dictionaries code up many mistrokes
to the same word as an attempt to satisfy this requirement. This
bloats up the dictionary, in my opinion. Emitting a mini-chart on
mis-translates gives the client a reasonable chance of determining
what was intended.

For training purposes, I prefer a "cannonical" dictionary, where only
the best stroke according to the theory translates, and any others are
flagged for reveiw. On the other hand, it is expecting too much for
clients to learn all the stroke substitutions on a misstroke, so the
compressed paper-tape format for mistranslates /TKUZ/-PBT/ quite meet
the bill (doesn't). However, the mini-chart form is definitly more
readable, even if it .d.._u'...z ....__'.n.t quite match what the
client expects; but the result is still more readable since most of
the common theory sound substitutions are in place.

I also find that casting the dictionary in this format flags theory
inconsistencies. The requirement is that the substutions apply
universally -- no dictionary is permitted for a mini-chart. --Lee

[Mirabai Knight]

I really like the example sentences in both these lessons, though I'm afraid I'm still having a bit of trouble following the intricacies of your ascii steno chart. I think it might come down to different brain patterns latching onto this or that pedagogical device and not being able to grok certain others; I'm sure it's very helpful to some people. Thanks for posting!

[leeo]

For those of you who get the e-mail digest, please forgive me for
resubmitting this series yet again. I spent much of my time over the
weekend editing and reworking the translation of the first chapter of
_Gregg Shorthand Manual, Simplified, Second Edition_. I have a
program that reads in the "mini charts" and writes the equivalent
"compressed paper tape", which also functions as a well-formed mini
chart verifier. There were many corrections. I added some sections
to present the material in a less haphazard way. Thanks Mirabai for
the encouragement, please check out Lesson 6, which includes a better
explanation of my unorthodox approach. I would invite a cross-
translation into the Plover Dictionary Theory. Unfortunately, being
unfamiliar with it, I would not be the best for this task. --Lee

It occurred to me that an established method might be useful for
introducing a machine shorthand theory. Having first learned a pen-
based system from _Gregg Shorthand Manual, Simplified, Second
Edition_, copyright 1955 by the McGraw Hill Book Company, Inc;
Copyright 1949 by The Gregg Publishing Company, I thought that I'd
translate the method across. I am fond of the textbook for its lucid
and concise presentation. The orders of letters introduced are
somewhat haphazard for machine shorthand since they are more natural
for Gregg shorthand. Thus the consonant sounds will be presented
unlike any machine shorthand method I've encountered -- the "natural"
key names are not introduced at first. This makes for an interesting
or perhaps frustrating "alternative" approach. I recommend practicing
on my "keyboard map" program which, if set to "mini-chart form",
immediately translates the strokes as described into the mini-chart
form used throughout.

In Gregg shorthand, the same form is used for a consonant preceding a
vowel as well as a consonant following a vowel, but with machine
shorthand, this is not the case. Also Gregg Simplified does not
distinguish long and short vowel sounds but machine shorthand does.
For this reason I have to simultaneously introduce both forms, and the
introduction of the consonant sounds and their positions may at first
seem arbitrary and strange. I have to beg the reader's patience here;
just persevere and learn the positions by route at first. The mini-
chart form suggests which keys are pressed for each form.

I'll continue translating lessons as time permits unless one of the
copyright holders either for the Gregg Shorthand or the Phoenix Theory
complains. It is my understanding that Gregg Shorthand, especially
the older Simplified version, being published over sixty years ago, as
fallen into being a lost art.