Sound vs. Spelling

• If someone offered you a dollar for either each word you can hear OR each word you can spell correctly, which would you choose?
• Which would make you the best shorthand writer: Realtime translation of each word you can hear OR of each word you can spell?

Question: I thought shorthand was written by sound. Why do you have to know how words are spelled before you can write them in shorthand for realtime translation?

Answer: Historically, shorthand theories have required that shorthand strokes be conformed to vowel spelling. For example, verse = VERS, worse = WORS, curse = KURS, first = FIRST. The vowel sounds are identical; but shorthand strokes follow the vowel spelling.

Question: Why is that a problem? Everyone knows how to spell those words. Are you saying writers of electronic shorthand shouldn’t have to know how to spell?

Answer: Any work product should reflect perfect spelling, of course. But there’s a big difference between:

• Preparing a perfectly-spelled transcript with the aid of an English language dictionary and a spell checker;
• Making sure your shorthand strokes reflect correct vowel spelling while writing machine shorthand at high speeds.

And there can be a huge difference between being a "good" speller—or even an "exceptional" speller—and being a "realtime" speller. Take the spelling quiz in the next column to see if you’re a realtime speller.

• Each word starts with the sound in. Is the correct spelling EN or IN?
• Each word ends in the sound unt or uns. Is the correct vowel spelling E or A?

Indicate your answers by making a mark through the correct vowel. Time yourself. Complete the exercises as quickly as you can.

If you completed both exercises in 45 seconds, you were making spelling decisions at a rate of approximately 70 words per minute. A text entry operator would have to make the same spelling decisions at 120-plus words per minute, a court reporter at 225-plus words per minute—both while continuing to hear, comprehend, and write additional words.

Were you 95% accurate? That’s the least that would be required on a realtime skills test—and would not be an acceptable level of accuracy on the job. It’s unreasonable to expect even expert spellers to conform shorthand strokes to vowel spelling at 120- to 225-plus wpm.

Consider the following when deciding whether it makes sense for realtime shorthand theories to be spelling dependent:

The English language is estimated at 400,000-800,000 words—of which the average adult has a recognition vocabulary of approximately 30-35,000 words and a use vocabulary of only approximately 10,000 words. The fact that you’ve never heard a word before probably means you can’t spell it—but it shouldn’t mean you can’t write it in shorthand.

A theory which requires knowledge of correct spelling before words can be written for realtime translation no longer meets the needs of real people, of the business community, or of the court reporting profession.

Question: Why doesn’t everyone just create their own translation dictionary? Then they could use shorthand outlines that match however they think words might be spelled.

Answer: If someone can’t remember the spelling of the word, it’s just as unlikely they’ll remember the spelling they decided to use in their dictionary entry. Also, creating a translation dictionary which is adequate for realtime writing is a very time-consuming, labor intensive project. Several thousand hours were spent creating the Phoenix translation dictionary and verifying that it is free of conflicts.

Question: If spelling dependence is such a big problem, why is Phoenix the only theory which has done something about it?

Answer: Before CAT (computer-aided translation of shorthand), it wasn’t that important. If we didn’t know the spelling, we just wrote the sound. It didn’t matter how we wrote a word as long as we could read it.

But for realtime electronic shorthand, if steno strokes don’t exactly match preprogrammed entries in a translation dictionary, they won’t translate. Technology has changed and the needs of the profession have changed.

Question: So how can we write realtime electronic shorthand without having to know how words are spelled?

Answer: In the Phoenix Theory we write by sound and we omit most of the indistinct vowel sounds.

Question: Whoa! Writing vowels by sound makes sense so we can write words we don’t know how to spell. But why omit vowels?

Answer: English has a vowel sound called the schwa vowel—sometimes aptly called the unvoiced vowel because it has no distinct sound. It sounds like an unstressed uh or, less frequently, a soft, unstressed ih. It’s the unstressed vowel sound heard in license, valance, grievous, parrot, rivet, palace, service, defendant, dependent, etc.

In kindergarten, we learned the vowels A, E, I, O, U—but nobody said anything about a schwa vowel. So it would be easy to assume it’s an uncommon sound, not found in many words, and couldn’t be that important to writing shorthand.

To the contrary! Per Funk & Wagnalls, the schwa vowel sound "occurs in most of the unstressed syllables in English speech." The schwa vowel sound is actually the most frequently used vowel in English speech. The majority of words of more than one syllable are pronounced with one or more schwa vowels.

Realistically, we must be able to write machine shorthand by sound, without a need to know how words are spelled in order to receive realtime translation. If we write by sound, we must have a key/key combination for the schwa vowel sound. But there’s no unused key or key combination on the keyboard!

After analyzing the problem, it was determined that the simplest and most effective solution is to omit most schwa vowels. (Phoenix Vowel-Omission Principle, Patent Pending.)

Question: Doesn’t omitting vowels make shorthand difficult to read?

Answer: As a matter of fact, it makes it easier to read. If a stroke doesn’t include a vowel, it’s verbalized with an uh sound: DAM/-J DAM uhj; VES/-L VES uhl. You’re less likely to mispronounce a word if you’re not influenced by seeing a vowel which doesn’t represent the correct pronunciation.

Question: Doesn’t omitting vowels create an awful lot of conflicts?

Answer: To the contrary! A major advantage of omitting vowels is that it automatically eliminates the vast majority of conflicts in word boundaries.

Homonyms are a nuisance. But they’re easy to recognize and they’re a nuisance we’ve learned to cope with.

The biggest area of conflict in machine shorthand, the one which has frustrated court reporters for the past 20-plus years and forced conventional theories to add dozens of rules which shorthand writers have to memorize and apply while writing, is the conflict in word boundaries. English uses the identical spelling for words, word beginnings, and word endings. For example: or, ordeal, candor; enclose, golden; in, insert, robin, etc. Frequently we even have one word spelled exactly the same as two words: carpet/car pet; bargain/bar gain, madam/mad am, etc.

Notice we said the spelling was identical. Other than homonyms, there are very few instances where pronunciation of a word or words causes conflicts. If we write shorthand by sound, and eliminate the schwa vowel sounds, we eliminate the majority of conflicts in word boundaries. It’s almost ironic that stubbornly clinging to an insistence on conforming steno strokes to vowel spelling is primarily responsible for creating the conflicts which we’ve been trying so hard—and sometimes so unsuccessfully—for 20 years to eliminate.

There are 12 shorthand theories presently approved by NCRA (National Court Reporters Association). But there are really only three distinct choices:

CONVENTIONAL THEORIES

Conventional theories are machine shorthand theories which were in existence before CAT (computer-aided translation of shorthand) and which have been modified in an effort to become sufficiently computer compatible to be viable for writing electronic shorthand for realtime translation.

Not only are conventional theories very spelling dependent, but they have two additional major drawbacks:

• In an effort to meet NCRA requirements, conventional theories have had to incorporate numerous rules to eliminate conflicts. These rules can add significantly to the difficulty of teaching, learning, and successfully writing shorthand.
• Many areas of conflict remain unresolved.

DIGITEXT

The Digitext theory was created post-CAT and was designed specifically for realtime writing and has successfully eliminated virtually all conflicts with a minimum number of "rules."

Digitext is unique in its use of an algorithmic translation logic to compare steno strokes to a chart of word/word part spellings. This unique translation system also makes Digitext unique in the degree of knowledge of word structure and spelling required for writing shorthand successfully for realtime translation. Additionally, Digitext writers must understand the functioning of the algorithmic logic in joining word parts (increments of spelling) together and be able to apply that understanding while writing shorthand.

The Digitext translation chart pairs steno strokes with specific spellings. Words are translated by joining increments of spelling together. Examples:

• The chart pairs the stroke YOUS with the spelling IOUS and the stroke YUS with the spelling IUS. The writer must know that the yus sound in genius is spelled IUS and must be stroked YUS but that the identical sound in ingenious is spelled IOUS and must be stroked YOUS.
• For some words, there may be little relationship between pronunciation and the shorthand strokes necessary to match the translation chart. E.g., region is pronounced REE jun, but must be stroked RAOEJ/YON to match increments of spelling; captious is pronounced CAP shus, but must be stroked KAPT/YOUS to match increments of spelling; pretentious is pronounced pree TEN shus, but must be stroked PRE/TENT/YOUS, to match increments of spelling; dissentient is pronounced di SEN shunt, but must be stroked DIS/SENT/YENT to match increments of spelling.

Digitext realtime translation is presently limited to keyboards containing the Digitext translation logic (Impact, Fon’iks Writer).

PHOENIX

Phoenix Theory was introduced in 1996. It was specifically designed for realtime translation of electronic shorthand and has successfully eliminated virtually all conflicts with a minimum of "rules."

Phoenix is the only realtime theory which:

• Has minimal spelling dependency;
• Uses a comprehensive Vowel-Omission Principle to facilitate lack of spelling dependency, simplify stroking, and avoid conflicts;
• Includes a conflict-distinction stroke for distinguishing homonyms.

Phoenix Theory can be written on any shorthand keyboard manufactured for writing the English language. The Phoenix translation dictionary (132,000-plus entries) has been electronically verified to be free of conflicts. It is available in Case CATalyst formats, and can be formatted for compatibility with other CAT software.

A good way to evaluate a theory—or to make a comparison between theories—is to examine the actual shorthand outlines. Is the theory operator-compatible, as well as computer-compatible? E.g.,

• Are steno strokes for identical sounds consistent?
• Do steno strokes represent a logical response to the sound of the word—or require rote memorization?
• Does it require the doubling of consonants, contrary to either correct spelling or pronunciation?
• Is steno for multi-stroke words unduly stroke intensive?

The following chart will help you in making a comparison. We’ve attempted to make the steno as "reader friendly" as possible for the benefit of people who do not read/write shorthand. The letter L, for example, will be shown as L rather than the actual steno stroke HR-. Theories may vary in key combinations for specific endings such as -TH, -RCH, etc. Those differences have little significance for your purposes here and won’t be demonstrated. Vowel sounds are also difficult for the non-writer/reader to make sense of and will be shown phonetically as:

• ay (pay) AI
• ee (see) EE
• eye (like) IY
• oh (show) OE
• yoo (you) YOO
• oo (suit) AOU
• aw (law) AU
• oy (boy) OI
• ow (how) OU
• double OO (food) OO

Shaded areas of the following table indicate spelling dependency