Shorthand on the QWERTY Keyboard, 1875-1917

“kn u rd ths? gr8!”

That might look like twenty-first-century mobile phone text messaging slang, of which it would in fact be a perfectly good example.  But that’s not what I’m using it here to illustrate.  Instead, I wrote it in one version of nineteenth-century “stenotypy,” following the guidelines put forward in a book published in 1888.  Stenotypy, in this sense, refers to the practice of taking shorthand on typewriters with ordinary QWERTY keyboards.  At a time when specialized stenographic keyboards hadn’t yet caught on, a number of people devised ingenious systems for doing this, anticipating many of the specific abbreviation strategies now associated with SMS language, but often elaborating them to a far greater extreme than anything current in “texting slang” today.  Their efforts have been overlooked and forgotten for the past hundred years.  Its hi tim 4 thr stry 2 b tld.


The early history of typewriters and “writing machines” combined two competing aspirations: on the one hand, to create attractive, legible documents more rapidly and conveniently than they could be written by hand or set in conventional type; and on the other hand, to get words down on paper as quickly as possible as an alternative to shorthand systems that relied on pen or pencil.  For the first scenario, a machine needed to be able to handle familiar alphabets and spellings and layouts.  For the second, however, there was no such expectation: the machine’s writing could take a very unconventional form, analogous to shorthand, as long as the operator could decipher it afterwards, the idea being that it might prove quicker and easier to strike one of several distinctive keys, or a combination of such keys, than it was to draw even the simplest equally distinctive mark on paper by hand.  The results could also be less ambiguous, since handwritten shorthand symbols existed within a continuum of shapes and positions and could easily be mistaken for one another, while typed characters were discrete units with clearly-bounded identities.

These two scenarios have long since come to be associated with two different standardized keyboard arrangements.

QWERTY keyboard from George Carl Mares, Art of Typewriting (1905); Stenotype keyboard from Ward Stone Ireland, Stenotypy (1914).

The first scenario has long been the domain of the QWERTY keyboard, introduced in the 1870s via the Remington typewriter.  One character gets typed on it at a time (with two options per key after the introduction of the shift key in 1878), and each keystroke causes the machine to advance one place in the current line.  Over the years, the arrangement of the outlying keys has changed significantly, but the positions of the basic letters have remained the same in most markets while undergoing only minor variations in others to reflect the needs of different languages (AZERTY in France, QWERTZ in Germany, etc.).  The QWERTY arrangement later found its way onto computer terminals and from thence onto smartphones, making it well-nigh ubiquitous today.  In practice, modern literacy pretty much demands a knowledge of it.

The keyboard now associated with professional stenography, known as the Stenotype, wasn’t introduced until a few decades later, in the mid-1910s.  It follows a logic very different from that of QWERTY.  Multiple keys can be struck on it at once, with (generally) the left hand covering initial consonants, the thumbs covering vowels, and the right hand covering final consonants.  When a “chord” is struck in this way, its characters all get typed at once in a single row and the machine automatically advances to the next line.  Words are encoded phonetically, ignoring their conventional spellings, and many speech sounds which the Latin alphabet would express with single letters are instead assigned to distinctive combinations of keys, such as PW on the left for initial B and PL on the right for final M.  While QWERTY is now routinely taught to elementary school children, Stenotype remains the purview of specialists, with proficiency acquired mainly on a need-to-know basis.  A small but enthusiastic subculture would like to change that; check out Plover if you’re curious.  Stenotype keyboards are uncommon outside of settings such as courtrooms, and relatively expensive besides, although some stenotypists substitute high-end QWERTY keyboards that allow “chording” with a subset of the keys mapped to the Stenotype layout (see

It follows from the foregoing that there was a period of around forty years, from the 1870s through the 1910s, when the QWERTY typewriter was commercially available but no specialized stenographic keyboard had yet emerged to compete successfully with it.  During that time, a number of people accordingly came up with stenographic systems for the QWERTY typewriter—systems that didn’t involve “chording” or, in most cases, any mechanical modifications whatsoever.  Sometimes individual typists devised their own personal systems or simply transferred strategies from handwritten shorthand to the keyboard, as suggested by the following piece in The Stenographer and Phonographic World for November 1915:

[S]tenotypewriting is fascinating, and there are many who are working it into practical shape.  There being, as yet, no reliable text-books on the subject, it is each one for himself.

Questioning a young typist (red of hair and blue of eyes) who had been “getting” the “old man” at a 210-word clip, as to the modus operandi, she brushed back a curl and laughed.

“Why, really, there’s nothing to it.  Just use your shorthand spelling.  Look here,” and with a flipping of dainty, white fingers she dashed off this—“ivntsnufrsmtm” (I have not seen you for some time).  “Anyone that knows shorthand spelling can read that.  And look here,” and the machine flashed out “SuM.”  That’s ‘consummation.’  It’s very simple.  Capitalize the ‘s’ for the prefix ‘con-‘, and also the ‘m’ for the suffix ‘-tion,’ and you have it.”

Notwithstanding this account, it wasn’t entirely “each one for himself” (or herself), since there were in fact several “text-books on the subject,” although whether these were “reliable” or not would have been a matter of opinion.  In the following post, I’ll summarize the rules of all the systems for which I’ve been able to find solid descriptions (and will list a few systems for which I haven’t).  Indeed, you’ll actually find enough information right here to teach yourself Byrne Stenotypewriting, Shorttyping, or Quinn Stenotypy to at least a good intermediate level if you like.  I’ve also provided authentic contemporaneous reading examples in several of the systems for you to try your hand at deciphering.  Word-game aficionados, gird yourselves!

Lane Stenotypy

Let’s begin our survey with a relatively simple system, the one outlined by George Lane in Hand-Book of Stenotypy; or, Shorthand for the Typewriter.  Being a Formulated System of Abbreviated Orthography for the Use of Typewriters and Others (1888).  Lane introduces his project as follows (p. v):

Occasionally, expert typewriters, those who find it inconvenient to leave their instrument, when called upon to take memoranda stenographically, resort to typewriting such in abbreviated orthography, and exercise their own ingenuity to express them so that they themselves, at least, may readily translate and transcribe afterward in full.  It is not known to the author whether any effort has been made to reduce “contracted orthography” to a system; if he is correctly informed there never has been, and the conception of such an undertaking is claimed as original, also the application of the title, if not the word itself—Stenotypy, or “shorthand typewriting.”  Being a “new thing,” he believes it susceptible of improvement, and to that end invites correspondence with typewriters who appreciate it.

In 1888, the word stenotypy was already a well-established term for the printing of shorthand word-forms in standard or “phonetic” character fonts, mostly to provide guidance in shorthand textbooks, and it had also been applied to pre-QWERTY mechanical approaches to stenography—see, for example, L. Danel-Duplan, La Sténotypie: procédé de sténographie mécanique (1864)—but we shouldn’t be too hard on Mr. Lane, since it’s tough to find a good name for something like this that hasn’t already been taken.  Lane proposes the following rules:

  • Leave out vowels wherever they can reasonably be inferred, e.g., learner → lrnr | missed → msd | arbitrary → rbtrr | enemy → nm | quail → qal.
  • Simplify diphthongs and drop silent or doubled letters, e.g., sea → se | ear → er | egg → eg | bought → bot | sway → swa | expugn → xpn.
  • Substitute j for ch as in choose, e.g., church → jrj.
  • Substitute k for hard c; f for ph, etc.; but retain characters from conventional orthography if they require no more keystrokes than their “phonetic” alternatives or are distinctive enough to help disambiguate, e.g., philosophy → flsf |  judge → jj | grotesque → grtsq | honor → hnr.
  • Shorten common endings, e.g., tion → tn | sion → sn | decision → dcsn | insidious → nsds.
  • Contract over (o’er) to or | overreach → orrj.
  • Use 00 for hundred and 000 for thousand.
  • Run together common phrases without spaces, e.g., of the → fth | ever so much → vrsmch.
  • Take advantage of existing abbreviations in common use (Lane provides a list starting on p. 97).
  • Use systematic abbreviations for months and days of week (e.g., M = March, Ma = May; full list provided on p. 56).
  • Substitute & for (a)nd, e.g., grand → gr& | amendment → m&mnt | andiron → &rn.
  • Substitute q for the sound of its name, e.g., secure → sqr | cubic → qbk | execution → xqtn.
  • Substitute numerals for the sounds of their names:
    • 2 = to, too, tu | tooth → 2th | to wit → 2wt | plenitude → pln2d | rectitude → rkt2d
    • 3 = thr | thrust → 3st | thrill → 3l | threat → 3t | through → 3u | throw → 3o
    • 4 = four, fore, fore | heretofore → hr24 | fortunate → 42n8 | foreign → 4n
    • 6 = sex | sextuple → 62pl | Essex → s6
    • 7 = sen, cen, seven | sevenfold → 7fld | sentiment → 7tmnt | century → 72r
    • 8 = ate, ait | relate → rl8 | emanate → mn8 | hateful → h8fl | gait → g8 | humiliate → hml8
    • 9 = nine, nign | leonine → lo9 | benign → b9 | feminine → fm9

These rules might seem fairly straightforward, but it’s not always clear how they should be applied to specific cases.  Thus, Lane often wavers between two or three alternative spellings for a given word: should it be ol or wl for owl, on or wn for one, frt or frut for fruit, bro or buro for bureau?  He isn’t sure, which suggests that he hadn’t yet put his proposed system through its paces himself at the time he wrote the book.  It’s also unclear whether he had much experience in trying to decipher its results.  Starting on page 59, he lists all the readings he can think of for each single letter and two-letter combination; bl, for example, could mean ball, bail, bale, bailey, able, ably, bawl, blay, abele, bell, belly, belle, belie, belay, bile, bill, billy, abole, abolla, boil, bowl, bole, boll, blow, below, bellow, bull, bul, bully, billow, billowy, buhl, boule, blue, blew, bluey, or by-law.  That’s a lot of ambiguity to swallow.  Of course, some ambiguity might be acceptable in a system intended only as a mnemonic aid—say, for taking down public speeches to be transcribed the next day while they were still fresh in someone’s memory.  Even so, Lane doesn’t dare provide any sample texts encoded using his system for his readers to try to figure out on their own.  The closest thing he gives us is some interlinear “pronouncing tests” starting on page 94, alternating between lines of stenotypy and equivalent lines of plaintext.  But just in case you’re looking for a challenge, I’ve copied out only the lines in stenotypy here for you to try to decipher.  Can you work out what they’re supposed to mean?  If you get stumped, you’ll find the solutions at the end of this post.

A lnnt prnt skd hs dtfl dotr wht id hd pssd hr prsmbl dkrs ant 2 akst th sklmstr wth sj a vrulnt trad f ralr  Th nqr dmnstrtd tht th erdt knsrvtr f pdgg wh hd bn hr nthsastk kojutr n orthp hd pkntl fo& flt wth th mtrns prnnctn & sh thr4 fl n2 a vhmnt pssn  Wth svrn dsdn sh objrgtd th unwr pdnt & ntrdcd hr nvntr f rvltng vkbls n gruls dnnctn f hs odcs drsvns  H hst& wth knsmt suvt 2 sfn hr xtrrdnr wrth bt fo& hr nrvtd ftr th sbsidnc f hr xhstv frt  Hr dcl ntr rsmd ts wntd kmns whn he ludd 2 hr ludkrs ennctn f th wrd suf & dsgntd hr grevs rr n th elsn f th ntl spr8 n th wrd wrf

N adtn 2 th flong abrvtns n kmn us thr r 00s f othrs tht r rkgnzd wthn th sfr f thr utlt blngng 2 th vrs brnjs f bsns & mnfctr  Evr rlrd stmshp xprs r othr krng kmpn pblk & prvt socatns & socits hv thr spcl abrvtd ttls wj t wld b qt mprktkbl 2 ntrdc n ths lst  Ths gvn r 4 rfrnc onl & hv n prtklr knktn wth th sstm f Stntp prvsl dsksd  Thr r lso a lrg nmbr f sns mblmtkl & othrws lstrtv f kmkl mdkl grmtkl kmcl tpgrfkl btnkl strnmkl & mslns sbjkts  Th l sh th frts tht hv bn md 2 rdc th prntd r wrtn rprsnttns f fkts & ids 2 th smlst kmps 4 thr sgnfktns w shl hv 2 rfr th rdr 2 th splmnt f Wrcstrs nbrdgd dktnr 2 wj a lrg prtn f th flong lst s krdtd


Byrne Stenotypewriting

If George Lane’s stenotypy gives the impression of a system that hasn’t been fully thought through, Byrne Stenotypewriting (1915) is another story entirely.  Its author, Henry Edward Byrne, was President of Tyler Commercial College in Tyler, Texas, which is said to have been one of the country’s largest business training schools.  The advertisement reproduced above, which appeared in the June 1916 issue of Popular Mechanics, claims that a number of institutions were then teaching Stenotypewriting, including high schools, YMCAs, and seventeen colleges in Chicago.  Another advertisement from the Stenographer and Phonographic World for January 1917, shown further below at right, goes even further, boasting that “hundreds of our large schools are now teaching it” and that more than five hundred teachers had signed up for a correspondence course about it.  Byrne’s system is unusual for making use of a distinction between characters struck “light” and “hard” on the typewriter, and its rules are as follows:

  • Letters and numbers can be used to represent the sounds of their names: see → c | seen → cn | be → b | for → 4 | before → b4 | decay → dk | you → u (also used for your, yours).
  • “Silent” letters are dropped, and short vowels can also be dropped, e.g., high → hi | rated → ratd | engine → ngn.  If a short vowel separating two consonants is dropped, the two remaining consonants can be further abbreviated as though they appeared side by side.
  • c is replaced by s or k.
  • u is dropped after q.
  • The “silent” parts of vowel pairs are dropped, e.g., beef → bef | eat → et.
  • y as a vowel is replaced by i (e.g., boy → boi), while y itself can represent ry, ly, ty, dy, or ity, as well as the consonant y.
  • th is replaced by a comma, which can also represent the word the.
  • wh is replaced by a hyphen ().
  • ch and sh are replaced by a semicolon (;).
  • ion, tion, cian, sian, tious, cious, cial, etc., are replaced by x.
  • sw, tw, and dw are replaced by x struck “heavy.”
  • sp and st are replaced by ¢.
  • ive and tive are replaced by ¢ struck “heavy.”
  • sk and sq are replaced by ½  thus, success → skss → ½ss.
  • ment is replaced by ½ struck “heavy.”
  • Some basic word signs are 2 = have | 3 = I, if | / = of | z = and | $ = dollar(s).
  • Words are ordinarily run together without spaces, and sentences are separated by periods (.), with no spacing.
  • However, spaces are inserted before and after any letters used as initials and abbreviations, as well as any numerals used to represent numbers (if a numeral is preceded by $, no preceding space is necessary).
  • The numerals 5-9 otherwise represent combinations of t, d, s, dd, td, or th with different vowels (5 for a, 6 for e, 7 for i, 8 for o, 9 for u).  If the numeral is struck with ordinary force, it means the consonant follows the vowel; thus, it → 7.  If it’s struck “heavy,” it means the consonant precedes the vowel; thus, ti → 7If a vowel is both preceded and followed by t, d, s, dd, td, or th, it’s treated like did → d7.  One possible reading of 8 “heavy” is to, which I suppose is why Byrne repurposed 2 for have.
  • If any lower-case letter is struck “heavy,” this means it’s followed by m, n, l or r, e.g., ham → ha | barn → ba| play pa | Mr → m | perfect → pfkt.
  • A lower-case vowel struck “heavy” can also indicate that it forms a two-letter word preceded by m, n, or l (e.g., my  i | no → o).
  • A lower-case consonant struck “heavy” can also indicate that it’s followed by s at the end of a word, e.g., rubs → rub.
  • c can represent com, con, coun, cor, col, etc. if struck with ordinary force (e.g., compromise → cpm7); or ng, ing, ding, ling, thing, etc. if struck “heavy” (e.g., roaring → roc).
  • 0 (zero) represents initial re, res, and apparently some other similar combinations.
  • Final s for plurals or possessives can be dropped whenever apparent from context.
  • Other letters are also dropped if they can be inferred from context, e.g., d in told → tol | find → fin | t in must → mus | b in begin → gin → gi | g in length → lenth → le,.
  • Unstressed initial vowels other than u are dropped, e.g., omit → mit → m7.
  • “When the context of a sentence will not make the word sufficiently clear to tell it from some other word when written with an ordinary outline, make the outline sufficiently full to make clear the word desired” (p. 23).

Examples (intermediate forms are shown as given in Byrne’s text, with spellings altered but before compression):

  • “He is a nice boy” → He is a nis boi → he7an7boi
  • “I see a big load of wood at the gate” → I see a big lod of wud at the gat → 3cabigl8/w95,g5
  • “I will be home today” I wil be hom toda → 3wibho85
  • “I was handling something, I knew not what, when it began flaming, roaring and popping and the smoke went rolling and winding in the air” → I was hanling sumthing I nu not what when it begn flaming, roring and poping and the smok wen roling and winding in the air. → 3w5hacsuc3nun8-5-e7bgfacroczpopcz,sokweroczwici,ai.
  • “Stop the ball of twine from swinging in the spider web.” Stop the bal of twin from swinging in the spider web. → ¢op,ba/xifoxigci,¢7rweb.
  • “The marshal was positive in his motive for slapping the fashionable fellow that lives in a mansion.  You should use caution when you are at the station, especially when there is a train in motion.” →The marshal was postiv in his motiv for slaping the fashonbl felo that livs in a mansion.  You shud us caution when you ar at the station specialy when ther is a tran in motion. → ,maxw5p8¢ih7mo¢4sapc,faxbfeo,5liviamax.u;99kaux-eua5,¢ax¢exy-e,e7ataimox.

Like Lane’s stenotypy, Byrne’s system introduces its share of ambiguities.  For example, 3wibho85 could be read “if wine be whole soda” as well as “I will be home today.”  Sorting out the intended meaning can feel a bit like doing a word puzzle.  “It is as much of an art to learn to read stenotypewriting as it is to write it,” Byrne acknowledges (p. 25).  If you’d like to try your own hand at this art today, a pair of “reading exercises” on pages 18 and 31 offer a prime opportunity, since Byrne originally published them without solutions.  I can confirm that it’s possible to decipher them based on the rules laid out above (my own solutions are given at the end of this post).  That said, I believe there’s one single typographical error in one of the two examples, but I won’t tell you what it is or where it is.  Good luck!



Byrne’s system also contains some further strategies for compression that aren’t needed to solve the puzzles above and might be considered more optionalif the system is workable, it’s probably workable without them.  I’ll outline the remaining strategies here, but without giving a complete list of character values as I have above.

  • Proper names begin with a capital letter, but capitalization can also be used to represent “prefixes,” e.g., A = accom, accoun | D = discom, discon, discoun, discor | E = equi, equiv.  Capitals struck “heavy” represent different prefixes, e.g., A = ambi, amphi | D = dissatis | E = ecom, encoun, encour | S = circum.  Some prefixes are also represented by consonant pairs, e.g., AD = adminis | ID = indis, indus | SS = substan.  A more exhaustive list appears on p. 43.
  • Capitalization can also be used to abbreviate the ends of words, including but not limited to “suffixes,” e.g., select → cL | provoke → poV | distinguished → d7T | circumstantial → SS | communication → cuK | irrecognizable → IN | rendition → 0D.  List on p. 45.
  • Common phrases may be reduced to strings made up of one character per word, with spacing before and after the string, e.g., dear sir → ds | your letter of the → ul/, | in reply to → irt | we are glad to know → wag8n.  Further examples on p. 39.
  • Very common phrases may be reduced to two characters, one capitalized to represent the first word, one lower-case to represent the last word, with a space following, e.g., hoping to hear from you soon → Hs | we quote you a price of → W/.  More examples on p. 40.
  • In court reporting, common phrases can be abbreviated in yet other ways, e.g., What is your name? → n.. | Are you acquainted with the defendant? → qd.. | Being duly sworn and examined on her own behalf testified as follows → sehobtf.  Two periods mark a question, and three periods mark a repetition. More examples on pp. 47ff.
  • Mistakes can be struck out by backing up and typing /.  Longer corrections, e.g., in dictation, are preceded and followed by //.
  • Byrne had designed his system with one particular keyboard arrangement in mind (shown at right), but it could be adapted to other keyboard arrangements with only slight modifications.
  • For example, someone using a typewriter without keys for ¢, ½, or 0 (zero) was instead advised to use:
    • q for sk and sq.
    • q “heavy” for sp and st.
    • r for re, res, etc.
    • v for ive and tive.
    • mm for ment.
  • A “double shift” keyboard required some further substitutions (p. 64).
  • On the other hand, the operator could take advantage of any extra characters that might be available to form additional word signs, perhaps with a light-heavy distinction or in various combinations with each other.
  • Byrne had also taken out a patent on a “Line Shift” mechanism designed to let the operator conveniently type characters a little below (or, on some models, above) the usual line at will, the idea being to use these line-shifted characters as distinct word and phrase signs.  He provides an “illustrative list” of signs on page 34, in which a line-shifted c represents Chicago, a line-shifted v represents ever or every, and so forth; but he seems to assume operators with Line Shift machines will define their own customized signs and tacitly avoids making the Line Shift essential to his basic system so as not to exclude the many users of ordinary typewriters.
  • Here’s a sample text embodying some of the more advanced features of Byrne’s system, including the Line Shift:

Quinn Stenotypy

The next system we’ll be examining, Stenotypy: or Shorthand by the Typewriter (2nd Ed., 1895), was the work of the Rev. Denis Alphonsus Quinn of Providence, Rhode Island, who is also known for writing a first-hand account of the yellow fever epidemics of the 1870s in Memphis, Tennessee, where he’d been a Catholic missionary.  His religious background plainly informed his choices about vocabulary to abbreviate (God, Jesus, Lord, Christ, etc.).

Quinn’s point of departure was the following observation (p. 10): “The chief impediment to fast typewriting is the frequent use of the spacer or space key, which requires a tap of the finger just like a letter of the alphabet, and is consequently very tedious where words of one syllable occur.  The system of stenotypy is based on the principle eliminating, or getting rid of space whenever this economy does not cause confusion in the reading of the subject matter.”  As this quotation suggests, Quinn’s treatment of the space bar is the most distinctive aspect of his system.  By reducing the usual need for a space between words, he frees up spacing to serve other purposes.

Two characters represent punctuation marks in Quinn’s stenotypy if they have a space before them, but prefixes if they don’t have a space before them:

” = com-, con-, cor-, col-
; = pri-, pre-, pro-, pru-

Some other characters are used as prefixes if preceded by a space but as suffixes if not preceded by a space; and when they’re used as suffixes they imply a word break, so there’s no space inserted after them either:

2 = contra-, contri-, contro-, counter-
3 = enter-, inter-, intro-, intru- | -try, -ary, -ory
4 = magna-, magne-, magni- | -fore, -dom
5 = trans- | -ive, -ife, -ful
6 = accom-, accum-, accoun- | -ion, -sion, -self
7 = encom-, incon-, encoun-, incog- | -able, -ible, -ment
8 = recoun-, recog- | -ate, -ity, -tude
9 = retra-, retre-, retri-, retro- | -ace, -acy, -ice
% = circum-, sub-, suf-, sur- | -ony, -mony
# = -ember, -gress
/ = -ake, -ike, -oke

Lower-case letters of the alphabet generally have their expected phonetic values.  However, upper-case consonants represent capital letters if preceded by a space, but entire words or suffixes if not preceded by a space.  Once again, the latter scenario can be taken to imply a following word break, so there’s no space inserted afterwards either (although in practice Quinn does sometimes add characters after them, e.g., Mn = amen).

B = be, by, but
C = say, see, sea
D = day, die, do
F = if, of, off
G = -age | God, go, age
H = he, have, him
J = -ness, -ship | Jesus, Jew
K = can, no, know
L = Lord, all, will
M = may, me, my, am
N = in, on, under
P = up, upon, people
Q = -head, -hood | queen, quack, quick
R = are, or, our, Mr.
S = she, shall, so
T = that, it, out
U = you, ye
V = very, every, over, ever
W = we, who, was, with
X = -oxy | Christ
Y = they, thee, thy, thou
Z = as, is, us

The following characters behave similarly:

& = -and | a, an, and, one
‘ = the
, = -ing, -s
– = -ly, -less
: = -us, -ous, -ious
@ = -ite, -ote, -ute, -ide, -ode, -ude | at
_ = -ward, -part
¢ = -ant, -ent, -end | cent(s)
( = th
) = sh
$ = ch | dollar, each, church

Meanwhile, upper-case vowels simply represent long vowels, while lower-case vowels represent short vowels (or, in combination, diphthongs).  Upper-case vowels can also represent capital vowels, but Quinn advises that there’s no need to space before them in this circumstance, as there is with consonants.  Small or short vowels “are seldom used in words of more than one syllable and never when silent” (p. 11).

Spaces for word breaks are also omitted:

  • Before a small vowel at the beginning of a word, e.g., he can’t irritate me → HKtirt8M
  • Between two identical small letters, e.g., short time )ttIm
  • After an s following a “logogram” or single-character word sign, e.g., The house was ours until yesterday ‘housWRsntl ystrD

On the other hand, spacing before and after certain lower-case letters forms the following abbreviations:

c = catholic, character
d = doctor
n = notwithstanding, nevertheless
o = circumstance, world
p = particular, peculiar
x = extra, extraordinary

Quinn also spaces before and after numerals used to represent numbers rather than prefixes or suffixes.

Finally, he puts forward a general rule (p. 11) to complement the others: “All consonants, vowels, and figures, such as B, C, I, K, T, 2, 4, 8, etc., stand for the words they phonetically express, C for see or sea, I for eye or I, T for tea, 2 for too or two, 8 for ate or eight, etc.  IOU&$&Ahlf means I owe you a dollar and a half.”

Here are a few examples of longer expressions in Quinn’s stenotypy:

  • You may bring me my chair, for I am tired of walking. UMbr,MM$r4IMtrdFwlk,. 
  • He who has seen the Mediterranean has seen the largest of seas. → HWhsCn’ Mdtrnn hsCn’lrgstFCs. 
  • You were all redeemed by the blood of Jesus Christ, our Lord and Saviour. UwrLrdmdB’bldFJXRL&svr.
  • The Lord’s Prayer: Rf(rWRtNhvn hlwdBYnmYk,4”YLBdnNr(ZTZNhvn g5Z(sDRD-brd&4g5ZRtrspsZW4g5(sWtrsps gnstZ&lEdZntN2tmpt6BdlvrZfrmEvl. Mn.

Up for another challenge?  If so, now’s the time for you try to decipher excerpts from two of Quinn’s “reading exercises” (from pages 20 and 17).  Note that the 0 in the third line of the first example is probably a typographical error for o.  The solutions, as provided by Quinn himself, are given at the end of this post.

r$&rAr wr’gmsSwOr
&&br@gld r,Nhr w&SbOr
BOhr bUtyWfrB0nd
hr sprkl,gms&snOwIt w&.

n(,mr5-tsts’Utl8F)rt&r@t,(n’rpd8Ww$Ths grwnN2Ze dr,’pst fUyrs.  50 yrsG&skld)rt&r@trWrAr&’rtWlkdPZ&srtFms3&S4md7Ftn7Z2dtrLB’mst crg:frm tempt,2f(mTs ms3s.


One of the few things I know about John Ira Brant, the author of Shorttyping (1914), is that he also wrote a utopian novel, The New Regime: A.D. 2202 (1909), in which he imagines a future world where all industry has been put under collective management and the savings from lack of competition has created abundance and leisure for all, but racially segregated hotels still exist (see p. 116).  Brant’s World War One draft registration card lists him as a “publisher” employed by the Blue Star Publishing Company that put out Shorttyping, and other sourcescensuses and suchsimilarly point to him working in the publishing industry for most of his life.  One advertisement, reproduced further below on the right, claims that Shorttyping had “gone thru the acid test” and that its “lessons are the finished product of actual teaching.”  Okay, that’s enough of a preface; let’s get down to the rules.

  • As usual, letters can be used to represent the sounds of their names: be → b | why → y | see → c | seen → cn | beat → bt | decay → dk.
  • As usual, “silent” letters are dropped, e.g., league → leag.
  • Beyond this, Brant’s system places a lot of weight on “precedence,” a hierarchy of possible readings for each character or combination of characters through which the reader is supposed to progress methodically until hitting upon one that makes sense, rather than puzzling out ambiguous cases from an assortment of equally likely alternatives.
  • In reading a word made up of two adjacent consonants, for example, the stated rule is first to try saying the names of both letters (e.g., xl = excel), then the name of the first letter (e.g., bn = bean), and then the name of the second letter (e.g., fr = far), going with whichever of these first yields a recognizable word.  If none of them does, then we’re supposed to keep inserting vowels until we get an intelligible result, in the following order: a as in take, a as in father, a as in all, a as in had, a as in hair, e as in eat, e as in yet, i as in ivy, i as in lid, o as in only, o as in on, u as in unite, u as in up, oi as in oil, ou as in out, oo as in food, and oo as in foot.  In typing, this approach would seem to require typists not only to think of the word they mean to type, but also to consider whether the same characters could form any other words of higher precedence, e.g., fad would have lower precedence than fade for the combination fd (because a as in take comes before a as in had), but tad would have the highest priority for td (since there’s no such word as tade).  The short form of each word would thus hinge largely on whether other competing words exist or not.
  • Spaces are inserted between words, such that it’s apparent whether a character comes at the beginning of a word, in the middle of a word, or at the end of the word, or whether a single character comprises a whole word.  This is important for what follows.
  • The figures (2 3 4 5 6 7 8 9 – ; , /) are each associated with one selection of syllables at the beginning or in the middle of a word, with another selection of syllables at the end of a word, and with one or more whole words.  In most cases, the syllabic values specify the consonants but not the vowels, which can vary and have been marked with ~ representing a generic vowel.  The readings given below represent (1) beginnings or middles of words, (2) ends of words, and (3) whole words shown in boldface, with pipes (|) punctuating these three categories.

2 = d~s, b~l  | b~l, ~b~l, ~mb~l | able, am
3 = com, con, cog | ~ct | can
4 =
k~r | ~nt, m~nt | other
5 = br~ | ~ns, ~nce, m~nt | that
6 = m~ss | ~ny, ~my, ney | any
7 = p~r | p~r | much
8 = pr~ | sh~n, san, zan, zhan, ~shun | such
9 = sh~ | ~sh, shall, shell, ential | shall
– = tr~, tar, ther | t~r, ther | there, their
; = st~ | ~st | state
, = sp~ | ~sp | has
/ = th~ | ~ng, ~nk, ath | this, thing
(final, in frasing)

  • Thus, 22 = disable | 7t = part | ac8 = action | oc8 = ocean | 8t3 = protect | 6/ = missing | 37/ = comparing | ;8 = station.
  • Advertisement from the Business Educator (September 1917)

    Readings are listed above within their categories in order of precedence, and syllabic values are generally supposed to be tried with the vowels a, e, i, o, u in that sequence.  In exceptional cases, syllabic values ordinarily associated with beginnings and middles of words can also be applied to ends of words, and vice versa, e.g., anticipate  4cpt (treating 4 as though it ended a word); dismiss → 26 (treating 6 as though it were at the beginning or in the middle of a word).  After all, Brant’s goal is not to distinguish rigidly between prefixes and suffixes, but only to make each symbol as practically useful as possible.

  • Vowels can be omitted whenever this doesn’t create ambiguity, e.g., confide → 3fd (not 3fid| dining-table → dn/t2 (not din/t2).  And according to the above rules, they can also be omitted whenever there’s no possible reading with a vowel of greater precedence.  Thus, prepare can be typed 87, but proper needs to be typed 8o7.
  • A vowel may also need to be typed to invoke an “ending” reading, e.g., ash → a9.
  • If a vowel is omitted before the sound k, implies a preceding long vowel, and implies a preceding short vowel; thus lk = lake and lc = lack.
  • The sound-based rules of precedence can, and should, be waived whenever they conflict with the goal of reducing the number of characters needed to type very common words: “The assigning of short forms to two or more words that enter into competition for the same syllabic character is, in fact, not a question of adhering to any inflexible rule, but rather a matter of the person who is using the system assigning the shortest combination to the word which he uses most frequently” (pp. 6-7).  By the same token, the choice between alternatives such as c and k might sometimes hinge on which is faster to type.
  • The characater i can be used for short or long i, but y as a vowel can only be used for short i, so it’s preferred whenever it helps to disambiguate short i from long i, e.g., lik = like, but lyc = lick.
  • In addition to their usual “phonetic” sounds and their names (as with dk = decay), most lower-case letters have additional sounds and/or words assigned to them, singly, doubled, and sometimes tripled; and single letters are also assigned whole-word values, shown below in boldface.

a = initial an or am | a, an
aa = crat
b = ba, bay | be, bee
bb = bant, bent
c = ic, ack | sea, see
cc =
initial scra, final cracy
d = day, ade, date, diate | day, and
dd = hold
e = the | the
f = for, full | for
ff = efficient,
or final ficient, fulness, fant, fent
(never soft) = gay, gate | give
gg = agnify, ignify, gant
h = hay, tch | which
hh = ther
(alternative to)
i =
initial in or im; ight (alternative to it) | in, eye, I
j = adge, edge, ju
 as in judge | not
jj = judge
k = cou
as in counsel, ac, coi, cate, quate | know, no
kk = final cific, lific, tific, entific, ratic, matic, cratic
l = lay, law | law, will
(in frasing)
ll = cler, final lessness
m = may, ember
| me, my, may
mm = member
mmm = memorandum
n = nay | on
nn = noon
o = oi, ou as in amount | of
oo = of our,
final owledge, ology, nology
p = pa, pay, pate, ship
(after n or r) | he
= quaccu, cute | to, too, two
qq = acular, nacular, icular, ticular, culiar
r = our, er
| are, our, hour
s = sub | is, us
ss = satis, ses
sss = sizes
t = th
(alternative to / ), ate | it
tt = trast, trist, trust, teristic
ttt =
the phrase trust it
u = oo as in look, initial um or un | you
uu = sult, zult
v = va, vay | have
vv = vant, vent
w = way, ow
 as in now | we, with
ww = spra
x = ax, li, ly | as
xx = change
y = short i | why, one
yy = dict, vict, divid
z = sla, ze, ize | was, use (in frasing)

Plus the following combinations:
ap = apt
au = ought
as in bought
by = body
cp = cept
ep = ept

  • Capital letters and several other symbols work more like the syllable figures described earlier, with the same three-way distinction drawn between values at the beginnings or middles of words, values at the ends of words, and whole-word or phrase values.

A = ab | bans, fans, gans, pans, tance, vans, zance | same
B = bla | bla, berg, bergh, bourg | but
C = sca | city, acity, icity, osity, uosity, scope | city, could
D = ad, od | der, dure, ded | did, deed
E = ed | ted | each
F = fra | fra, graph | from
G = gra, gar | gra, gar | against
H = har, char, cham | har, ham, han, hand | hear, here
I = il, ill, ig | ty, thy, aty, ity, dity, tity, anity, ice, vice | if
J = jar, jan, jarn | jar, jan, jarn | some
K = kal, kall | kal, kall, ical, dical, tical, itical, apital | upon, call
L = al, el, ul | lar, ality, ility, tility, bility, ability | all
M = mar, am, em | mar, gram, gramme | more
N = nar, an, en | nar | nor
O = ob, ord | ord, cord | who
P = pla, pan | pla, pan | had
Q = quar, cure | quar, cure, scure | after
R = ar, er, sar, sir | sar, sir, arity | or
S = as, es, ses | ses, as, es | so
T = out, et, it, ot, ut | tive, tant, tent, out | out
U = up | duce, dule, tude | up
V = var | ivity, tivity, vate, var | very
W = war | ward, word, war | where, wear, were
X = ext, man | man, ext | him, man
(final, in frasing)
Y = yar, sure | sure, tsure, ture, zure, yar, ion | year, sure, your
Z = sal, cel, cell | dize, lize, mize, nize, rize, size, thize, vize, wise, sal, sel, cell | sell, whose
” = ten, tem, trans | ten, tem, assable, sessible, sistible, ustible | time, ten
(in frasing), that the
# = stra, super, claim | stra, ster, straction, struction, claim | would, by the
$ = dra, contra, counter | dra, counter, sation, cession, zession, ization, lization, lition | many, of a, of an
% = tel, tell, recom, recon, recog, recor | tel, tell, b(e)ration, d(e)ration, g(e)ration, t(e)ration, tuation, stration, stition, stitution, poration, tiation | them, of the
_ = bate, unde, under, undou, indur, endure | bate, under, stand | under, under the
& = and, mand, accom, accomp | and, mand, imate, timate, estimate, ulate, tulate, tunate, uctuate, icipata, ipulate, itulate, porate | they, and the
‘ = fla, over, aver | fla, bation, dation, gation, kation, lation, mation, nation, pation, qution, ration, tation, vation | over, at the
( = fal, appra | fal, full | made, for the
) = pal, impar | pal, ample, imple, eous, ious, uous | make, from the
: = far, incom, incon, incog, incor | far, fy, ify, cify, dify, nify, tify, form | what, with the
. = bar, extra, exter | bar, ry, try, ary, dary, mary, rary, tary, trary, porary, endary, oundary, ordinary | whom, in the
? = val, enter, inter, intra | val, action, ection, iction, uction | been, to the

  • The vowels shown in the above list of syllables are defaults but can be “cancelled,” e.g., hiO = hired (hi+[o]rd); 3u& = communed (com-u-[a]nd); dnV = Denver (dn+v[a]r).
  • Whenever a given sequence of characters could have two or more legitimate meanings, the typist is supposed to assign the sequence to one of them and find some alternative way of typing the other in order to avoid ambiguity.  Brant (p. 11) gives the example of iff, which could be read either inefficient or infant; he opts to assign it to inefficient and to use the alternative if4 for infant.  This strategy assumes there will be multiple options available for encoding the same speech sounds and helps explain some apparent redundancies in the system.
  • The existence of word-signs would need to factor into precedence as well.  For example, since ) = make, that presumably frees up the combination mk for some word that would ordinarily have lower precedence, such as meek.
  • Resolution won’t always be possible at the single-word level.  For example, since 2 as a word sign is assigned to both able and am, the context of surrounding words would be required to disambiguate it.
  • The general rule is to insert spaces between words, but an exception is made for what Brant calls “frasing,” i.e., typing multiple characters representing a common phrase without spacing in between them, sometimes reducing words to their first letters (executor or administrator → eoa), sometimes combining word-signs or phrase-signs (on which → nh), sometimes using capital letters to show “principal” words (Chief Judge of the Court of Appeals → CJ%CA), and sometimes through a combination of these approaches.  Brant provides a long list of “frases” starting on page 68.
  • Brant doesn’t provide any extended reading exercises, but here are a few of the longest expressions he offers: It is not the same → t s j e A | Are you sure you can see him? → r u Y u 3 c X | We know he has been there for some time → w k p , ? – f J ” | The records of the corporation → e %ds % 4%


Phonoscribing (Fo)

The next system, “phonoscribing,” was also known as “Fo,” which I take it must be the word “phonoscribing” phonoscribed.  Its inventor was Colonel S. N. Stewart of Philadelphia, better known for having built a pontoon bridge across the Missouri River in Nebraska City.  He always seems to have gone by “S. N.,” to the point that I don’t know what the initials stood for, but I suppose that shouldn’t be a surprise coming from someone who took abbreviations as seriously as he did.  Instruction in phonoscribing was restricted to a single school in Philadelphia that offered—according to a circular quoted in the Brown & Holland Shorthand News for July 1883—to

teach shorthand, brief longhand, writing machines and phonoscribing (shorthand on the writing machine, as readable as common print, because it is common print), and teaches all that can be taught in other shorthand schools, and phonoscribing in addition, which no other schools are permitted to teach.  The manual of phonoscribing cannot be bought.  Phonoscribing is a system of syllabic strokes for the pen or writing machine, of which there are 12,000 machines in use in the United States.  The average shorthand writer may be able to write the skeleton outlines of 100 words per minute on special run of words.  One hundred and ten such words have been written in full on the machine by different writers.  You can see it done under the watch if you are skeptical.  Three hundred and forty words a minute have been written in skeleton outline on the machine on a special run of small words.

However tightly the teaching of phonoscribing might have been controlled, the Phonographic World for October 1885 published a detailed account of it which I copy below at length.  Note that in some material quoted from a circular, of is spelled v.  I’ve corrected a handful of obvious typographical errors (after all, you can always consult the original if you like), but I’ve tried to leave the extended passages in phonoscribing untouched, even when they seem to violate the rules of the system.

[Phonoscribing] consists of a system of abbreviated writing for the typewriter, invented by Col. S. N. Stewart, of Philadelphia, and is divided into two classes, or styles, the “Corresponding” and the “Reporting.”  We are indebted to Mr. B. W. Readshaw, 128 Bird Ave., Buffalo, N. Y., for circulars from which we are enabled to draft the following respecting this system:

The “Corresponding” style of Fo may be best comprehended by a careful observation of the following seven rules.  They are easily understood and their adoption by the operator will be found to greatly add to speed, without, we think, at all detracting from the legibility of his writing.  We give them as shown by the circular in hand:

RULE 1.  Omit short e except where it begins a word or accented syllable: tl, tell; ltr, letter; evr, ever.  But e is always omitted before x: xtra, extra.

RULE 2.  Omit useless letters: tho, though; ot, ought; hi, high.   In a few words the omitted letter must be replaced by a letter that will indicate the proper sound, as vu, view; thru, through; wa, weigh.

RULE 3.  Change ph to f wherever those letters have the sound v F, fotograph [sic], photograph.

RULE 4.  Change G to J wherever it has the sound v J, hj, hedge; rij, ridge; juj, judge.  D is silent before g, but cannot be omitted unless g is changed to j.

RULE 5.  The plural v words ending in y is formed by adding s, pony ponys.  Y is also retained in the comparative & superlative degree v adjectives, as holy, holyr, holyst, and in the past tense of verbs; huryd, hurried.

RULE 6.  Any vowel before nd, ng, nk, or nt, is omitted unless it begins a word or is preceded by another vowel: hnd, hand; sng, sing; bnk, bank; wnt, want.

RULE 7.  Omit all vowels from any unaccented syllable or any word v 1 syllable, which, in the singular has 4 or more consonants, thnk, think.


To furnish some idea of the saving to be had by the use of Phonoscribing even in the corresponding style we give one verse of a poem appearing in the circular before us, entitled “A tale of Philadelphia True Love,” from which it will be seen have been omitted 37 useless letters, with no sacrifice to legibility:

O she ws a maid v a lafng i
Who livd n a gart cold & hi,
& he ws a thrdbr whskrd bo
Who dwlt n a clr damp & lo


In further illustration of the system and in denunciation of English orthography, H. Gamble, the publisher, 912 Arch street, Philadelphia, says:

“Th comn splng is the silyst thng in histry.  It is a lng wa round.  It is unnaturl.  It hlds u bac.  It rqirs mntl efrt & divrts ur enrjys frm cmpozishn, bcaus one sound is rprsntd by 20 or mor difrnt ltrs or cmbnashns of ltrs.  Thr is no mor rezn fr uzng th splng of our ancstrs, thn fr usng thr clumsy tools, imprfct mashens & rud vehicls; ride on th cars.  If one of th most clebratd of our projntrs shd aper on one of th st toda undr th ratl-trap umbrla, whch he cnsidrd a modl of prfcshn, frsky juvnils wd salut hm wth dcayd frut; bt hz splng is infnitly wors thn hz umbrla.  Awa wth it.

“U wondr tht th majority shd hv rzistd th intrducshn of steam powr.  Our dcndnts wil wondr y we clung to our crookd splng & rzistd strat & honst orthografy.  Tha wil sa, ‘Our ancstrs semd to undrstnd som thngs, bt tha cd nt spl.’  Tha wil pic out sch wrds as dough, cough, through, tough, plough, hiccough, hough, ought, & smil ovr thm, & whn one shl sa, ‘th ough in ech of thoz wrds rprsntd a difrnt sound,’ othrs wil refuz to blev it.

“Se th holo hmbug u vnrat: u laf at Dr. Gregory’s,

‘Syrrh: —Eye seigh yew propeaux two introwdeauce inn ewer jolonel ay nue sissedem achph righting…..
‘Y reamaighn ewers troughly,
‘A louvor achph phachddh scghoelneaund.’

‘(SIR:—I se u propos to intrduc in ur Jurnl a nu systm of ritng, I rman, urs trly, a lovr of bad splng,) yt ech of th abov splngs & hundrds of othrs in Dr. Gregory’s ltr cn be jstifid by wrds in th comn splng—wrds in whch th same round-about combinashns of ltrs r uzd.  Yt ‘familr wth th face’ we clng to th old splng wht th same tnacity tht th Turk clngs to th woodn plow of hz grndfathr.”


So much for the “Corresponding” style.  In the “Reporting” style, the author proceeds to get down to business in earnest.  He starts out with 18 additional rules, twisting all the letters of which the typewriter is capable into every imaginable shape and size for the contraction of words and the representation of syllables.  He then gives a somewhat extended list of “word-signs” (something after the manner of phonography) and winds up with a “List of Syllabic Signs of Letters” as follows:

The apostrophe (‘) is tr, ‘m, term; m’, matter.

6 is st.  6v, stove; h6, hoist.

The left parenthesis is pr.  (1, pearl; p(, paper; p((, proper.

3 is th.  3o, though; f3, faith.

4 is fr.  4m, frame; r4, refer.

The hyphen is rt. -k, retake; b-, bright.

$ is tl.  $FN, telephone; r$, retail.

7 is sh.  7m, shame; r7, rush.

8 is at or ad (at the beginning of words).  8n, eighteen; 8l, addle.

9 is mr.  9l, moral; r9, rumor; 99, murmur.

; is tn.  ;1, tunnel; c;, cotton.

: is str.  :1, sterile; c:, coaster.

? is inter, enter, under, intro.  ?vn, intervene; ?;, entertain; ?tk, undertake; ?dc, introduce, w?, winter.

5 is fl.  5m, flame; j5, joyful.

The right parenthesis is pl.  )t, plight; r), ripple.

q is lt; but only used at the end of a word, dq (dlt) delight.

j indicates l preceded by a vowel. jr (alr) allure.

y is r preceded by a vowel. yv (arrive).

z is s preceded by a vowel.  zm (asm) assume.

The following phrases of frequent recurrence are represented as shown: zp, as soon as possible; zm, as a matter of course; mrn, more than; 8v, it is very necessary; uc, under the circumstances; 2g, to a great extent; ha, I have no doubt of it; 8p, at the present moment.  Such phraseograms also effect great savings in telegraphing.

A repetition or a contrast may be omitted after &: gn &, again and again; rnd &, around and around, gd &, good and bad; blc &, black and white.

We give the above, clipped from extended circulars, only to furnish an idea of Phonoscribing.  We cannot pretend to give the whole, but what is here presented will serve to explain sufficiently to shorthand writers the subject in hand—to keep them posted concerning what is going on in the world of which they are a part and to enable them to “keep up with the profession.”

The Corresponding Style seems pretty fully outlined here: it’s comparable to George Lane’s stenotypy, but more legible and less compressed.  The Reporting Style more closely resembles Brant Shorttyping, but it’s unclear how much of the total system is actually spelled out here.  The reference to its “eighteen rules” and “word signs” could point to the material summarized afterwards—there are, in fact, twenty stated “rules,” plus the “phrases of frequent recurrence” inserted between rules nineteen and twenty—but the author/editor seems to imply that these are separate things and downplays the completeness of the summary as a whole.

Universal Declaration of Human Rights

The first article of the Universal Declaration of Human Rights has become the standard test piece for constructed languages, so I decided to translate it into each of the historical systems of QWERTY typewriter shorthand for which I think I’ve found enough documentation.  I’ve tried to follow all the stated rules, and when those haven’t provided a clear answer, I’ve tried to find the same (or similar) words in contemporaneous sample texts to use as models.  Even so, I’ve sometimes had to fall back on guesswork (what’s the best way to type “one” in Byrne Stenotypewriting?), and I’d be surprised if I haven’t made at least a few out-and-out mistakes.  Still, I thought it would be worthwhile to simulate how each system would have handled one particular text for purposes of comparison and contrast.



All human beings are born free and equal in dignity and rights. They are endowed with reason and conscience and should act towards one another in a spirit of brotherhood.


Phonoscribing, Corresponding Style

Al humn bengs r born fre & eql n dignty & rits. They r endowd wth rezn & conshns & shd act towrds one anothr n a spirt v brothrhood.


Phonoscribing, Reporting Style (as summarized)

J humn bengs r born 4e & eql n dignty & -s. 3ey r endowd w3 rezn & con7ns & 7d act towrds one ano3r n a spi- v bro3rhood.


Lane Stenotypy

L hmn bngs r brn fr & ql n dgnt & rts Th r ndwd wth rzn & kncnc & shld kt twrds wn nthr n a sprt f brthrhd


Brant Shorttyping

L huX b/s r .n Fe d eql i dgnI d rits.  & r ndwd w r8 d 39ns d 9ud kt tWs y a4 ia ,rit o 5-hd.


Quinn Stenotypy

LhUmnB,sRbrn frE&EqlNdgn8&rIts. YRndwdWrEsn&”6s&)d ct t_s&&(rN&sptFbr(rQ.


Byrne Stenotypewriting

ahuabcabonfrezeqai7gnyzr7s.,iae8wdw70nzcxz;9akt8wads 1 a8rii7/br8eh8. 


Other Systems

To the best of my knowledge, the earliest published system of stenotypy for the QWERTY keyboard actually appeared in 1879, the year after Remington introduced its No. 2 typewriter, the first to come equipped with a shift key for upper-case and lower-case letters.  It was the work of Edward Fitch Underhill—the “Renaissance man of Siasconset,” as he’s called in the title of a recent biography by Margaret Moore Booker.  Among other things, Underhill was a stenographic reporter for the Surrogate’s Court of New York City, and his 144-page book on stenotypy is entitled Steno-Printing: a System of Abbreviation adapted to increase the speed of the type-writer and make it available for stenographic uses (New York: Evelyn T. Underhill, 1879).  Although WorldCat reports six library copies, none of them appears to have been digitized, so for the moment I’ll need to content myself with the following description of Underhill’s system given by James G. Petrie in his paper on “Reporting and Transcribing Machines,” Shorthand 1 (1881-82): 155-6.

I have carefully gone into this system of Mr. Underhill’s, and cannot help congratulating him on the skill he has displayed in adapting the machine to very rapid writing.  I can only here, of course, give a mere outline of his system, which is founded on Stenographic principles, and which could only have been worked out by a practical Stenographer.  The exercises commence with simple abbreviations, such as ab- for about, bcm for become, f- for for, fm for from, and so on.  Then prefixes and suffixes are treated in the following manner:—for and fore are represented by the letter “f,” as fgv- forgive; fgo- forego; like by lk- as lkwz- likewise; mnlk- manlike, &c.  Circum is represented by cc- as ccflx- circumflex.  As suffixes, we find aught represented by ot- as frot- fraught; noty- naughty.  Cl stands for clude, as xcl- exclude.  Then again arbitrary representations are made by employing the figures 2, 3, 4, 5, 6, 7, 8, and 9, and the punctuation marks, the comma, semicolon, hyphen, and exclamation point.  To each figure is assigned the representation of both a prefix and suffix; for instance, con and com are represented by the figure 4; as 4ps- compose; 4fn confine; while 4 as a suffix stands for rt- ; for example, rt 4- retort.  Then again 6 as a prefix stands for pre; for example, 6fr- prefer; 6vl prevail; and 6 as a suffix represents ble; hnr6- honourable; ps6- possible.  The prefixes and suffixes are combined to represent the appropriate adjuncts and syllables occurring in conjunction—for instance, accom is represented by the letter “a” joined to the figure 4- com-; as a4dt- accommodate; discom by ds added to 4; as ds4ps- discompose; while in some instances the larger portion even of whole words is represented by combined prefixes and suffixes; as 54f46- uncomfortable.

Underhill’s system resembled Brant Shorttyping in its general approach, but based on Petrie’s summary it seems not to have made any use of the distinction between upper-case and lower-case letters.  Had Underhill begun inventing it before the introduction of the Remington No. 2, when this distinction hadn’t been available?  Or did he consider it important to continue using capital letters to represent capital letters?  Or is Petrie’s description just incomplete?  I guess I’d need to get my hands on Underhill’s book to find out.

Another publication which I haven’t yet seen is Stenotypy: A System of Short-Hand for the Caligraph, Typewriter, or other Writing Machines (St. Paul, West Publishing Co., 1884), by Harcourt H. Horn, reporter of the Tenth Judicial District of Minnesota.  In this case, WorldCat lists a single copy held by the New York Public Library.  For now, I’ll have to let the following advertisement from 1886 for Horn’s book stand as a placeholder:

Stenotypy gives a speed equal to any machine shorthand.  It is simpler and less complicated than any other system; employs but few arbitrary signs, and takes the action of the fingers rather than that of the brain.  The operator on the writing machine, according to this skill thereon, starts with a large portion of the requirements for Stenotypy already in his possession.  It is progressive.  While no real use can be made of shorthand before the whole system is explained and learned, the beginner in Stenotypy can put his knowledge into operation from the start, and thus increase his speed day by day.  No teacher is necessary, and little, if any, assistance can be rendered by one.

THE ABBREVIATED STYLE, which provides for a few simple contractions easily learned and committed to memory, will enable the operator to attain a rate of speed of 80 to 100 words a minute,—which is generally fast enough for ordinary dictation and commercial purposes.  By means of THE CONTRACTED STYLE, a fair operator can soon acquire, without difficulty, a speed of 120 words a minute, while a rapid one can attain to 150 or more.

The Author’s Manual will be found to be a Complete Hand-book dealing fully with both styles of Stenotypy, and written expressly for self-instruction.  The whole system is thoroughly explained, and copious exercises, with keys to the same, are appended.

Google Books also shows an entry for something called Stenotypy: A Key to Rapid Type Writing (1890) by C. R. Needham, which isn’t viewable and is just one page in length.  The author is presumably the same Clement R. Needham who wrote A Synopsis of the Physiology of Articulation (London, 1859), subtitled “adapted to reporting by machinery, and for telegraphic purposes,” in which “Stenotypy; or, Printing by Points” involves a hypothetical ten-key machine, not a standard typewriter.  But the same book also outlines a system of spelling reform based on standard type.  So who knows what could actually be in that one-page document?

Meanwhile, other publications that might cover relevant systems seem even further out of reach.  One of them is a book self-published in 1881 by Frederick Glanville in Venice, New York, with the title—as given in a bibliography of shorthand—of Index Writing.  A series of Papers, or Treatises on Orthographic Stenography, or a scientific aid to correct spelling, combined with a concise system of short-hand.  Alphabetically evolved from the repeated and unrepeated letters in the words symbolized.  Diagrammatically illustrated by the phases or variations of the sun’s circle of illumination on a planet’s disc.  Designed to facilitate Telegraphing, Type-writing, Reporting, &c., and to supersede the tedious, if not pernicious practice of conning the letters of words in order to memorize them.  The sun is not only the great electrical light of the natural and material, but also of the mental and moral world.  And the methods of the handwriting of the ordinances of heaven are simple if not obvious.  As by the mighty luminary, the divine beneficence is diffused through all the earth, impressed on its harvests and expressed in its seasons; so also by it are we furnished with rudiments of writing and reading, as a direct and sure path to mental and moral improvement.  “Their lines are gone out into all the earth, and their words to the end of the world.”  Glanville’s book isn’t listed in WorldCat, and there appear to be no known copies of it extant anywhere, which is really too bad—it sounds like a hoot.

The same bibliography also contains a listing for Longhand shorthand; a system of brief longhand for general use, requiring but a few days for its thorough mastery (Valparaiso, Indiana, 1883).  Its author, M. G. Kimmel, is described as Professor of Phonography and Typewriting at Northern Indiana Normal School, now Valparaiso University, but the booklet—which originally sold for seventy-five cents and, we’re told, involved a system “adapted to the writing machine”—seems to have vanished without a trace.

It’s probably impossible to determine who the first person was to devise a personal system of stenography for the QWERTY keyboard, but J. D. Strachan writes as follows in his article “Machine Shorthand,” published in the proceedings of the 1906 annual meeting of the National Shorthand Reporters’ Association (at p. 154):

Among the first to experiment with the typewriter for court reporting was F. Deming, of Albany, official reporter of the Third Judicial District of the Supreme Court of New York, who in 1876 used an extension, easily fingered keyboard and a few simple abbreviations.  His machine was partly covered with glass to deaden the noise of operation.  He used the machine for about three years, until the old city hall was burned.   A large amount of reporting was done with it.  That was when the typewriter was in its infancy.

The reporter’s name was actually Philander Deming, but abbreviating “Philander” as “F” might have been a political choice rather than a mistake—a blow struck at the heart of orthographic tyranny.  As it happens, Deming had patented a modified typewriter keyboard layout for stenographic use in 1875 with all capital letters on the left and all lower-case letters on the right.  Instead of spacing between words—which were to be spelled phonetically—the first letter of each word was to be capitalized, and all other letters were to be lower-case.  Or the contrast could be between two other categories of letter form; the important thing was that they should contrast with each other.  Whether the keyboard itself counts as a member of the QWERTY family is debatable, but the substitution of capitalization for spacing, which was technically a separate claim in his patent, falls into very much the same tradition we’ve been considering.

Petrie provides more details:

Mr. Deming, an American Stenographer, has used the type-writer in Court for reporting the proceedings, and has adapted to it an ingenious system of Stenotypic reporting.  His machine is provided with a glass case, which renders the mechanism practically noiseless in working.  It is evident that were one to use even such a system of steno-printing as I have just been explaining [i.e., Underhill’s], they would fall far short of the requirements of Court reporting.  An average of two letters to each word would require to be struck, and a third touch would have to be made for the space between the words.  By omitting the space a third of the average time is saved, and the rate approaches that of a good system of Shorthand.  At the same time a continuous line of letters only would be very difficult to divide into words, and therefore the spacing bar of the type-writer has been connected with a dotter.  By the mere turn of a button, and while the initial letter of a word is being struck by one of the fingers, the dot is made over that letter by the thumb simultaneously striking the space bar.  The emphasis of the initial letter thus soon becomes a habit, and is done quite mechanically.

For instance, suppose we write “This is a new way to do it.”  It is turned out by the type-writer in the following fashion:—


Mr. Deming states that he has used the type-writer in Court for a week at a time, but he prefers to use it hour and hour alternately with the pen or pencil.  He affirms that he thus experiences considerable relief from the exhaustion caused by a long spell of writing Phonography.

The “dotter” was the subject of another Deming patent, filed in 1875 and received in 1876, in which he calls it a “marker” or “divider” and notes that it could be either connected to the space bar or added as an extra key.  Yet another patent from 1876 introduces an equivalent “divisor-bar” that could be placed wherever the operator could reach it most easily from any point on the keyboard.  With the introduction of the Remington No. 2 in 1878, of course, operators could simply have substituted the shift key for the dotter, alternating between capitals and lower-case letters as Deming had originally proposed: This is a new way to do it →ThsSANwWyTDT.  In spite of the fact that Deming focused so much of his energy on mechanical innovations, I still see him as trying mainly to convert the standard typewriter into a stenographic tool rather than to invent an entirely different machine from scratch, comparable to the Stenotype.  The Line Shift patented for use in Byrne Stenotypography is functionally equivalent to Deming’s dotter.

Could any of these systems still be useful today?

The most recent references I’ve found to anyone teaching a system covered in this post date from 1917.  Since then, anyone with a serious professional interest in machine stenography has presumably turned to the chorded Stenotype instead.*  But the demand for shorthand systems suited to the QWERTY keyboard lives on among people with less daunting speed requirements.  The best evidence for this is the very existence of specialized conventions for text messaging, the resemblance of which to the older conventions of stenotypy is downright uncanny—witness the following table from Wikipedia.

Initialisms, too—on the pattern of BFF or LOL—were as much at home in stenotypy as they now are in text messages.  One major difference is that the systems of stenotypy we’ve been surveying were purposefully designed and formally taught, whereas proficiency in “txting” is generally picked up through observation and imitation, and the systems themselves seem to have grown up organically.  But if such conventions serve a real practical purpose in boosting typing speed, then I see no reason why standardization, further elaboration, and formal training couldn’t also be beneficial today.  Maybe it’s time for a revival!

Closing thought: the critics who grumble about kids typing cul8r today might be shocked to learn that students a hundred years ago were being taught to do the same thing in business schools, and that it was then regarded as a valuable skill that would get them jobs and make them productive members of society.

Solutions to the Reading Puzzles

Lane’s Stenotypy:

  • A lenient parent asked his dutiful daughter what idea had possessed her presumably decorous aunt to accost the schoolmaster with such a virulent tirade of raillery.  The inquiry demonstrated that the erudite conservator of pedagogy, who had been her enthusiastic coadjutor in orthoepy, had piquantly found fault with the matron’s pronunciation, and she therefore flew into a vehement passion.  With sovereign disdain she objurgated the unwary pedant, and introduced her inventory of revolting vocables in garrulous denunciation of his audacious derisiveness.  He hastened with consummate suavity to soften her extraordinary wrath, but found her enervated after the subsidence of her exhaustive effort.  Her docile nature resumed its wonted calmness when he alluded to her ludicrous enunciation of the word “sough,” and designated her grievous error in the elision of the initial aspirate in the word “wharf.”
  • In addition to the following abbreviations in common use, there are hundreds of others that are recognized within the sphere of their utility, belonging to the various branches of business and manufacture.  Every railroad, steamship, express, or other carrying company, public and private associations and societies have their special abbreviated titles, which it would be quite impracticable to introduce in this list.  Those given are for reference only, and have no particular connection with the system of Stenotypy previously discussed.  There are also a large number of signs, emblematical and otherwise, illustrative of chemical, medical, grammatical, commercial, typographical, botanical, astronomical and miscellaneous subjects.  They all show the efforts that have been made to reduce the printed or written representations of facts and ideas to the smallest compass.  For their significations we shall have to refer the reader to the “Supplement” of Worcester’s Unabridged Dictionary, to which a large portion of the following list is credited.

Byrne Stenotypewriting:

  • He is a good boy.  Do you like the girl.  See the man with my ball.  Come and play with me.  He will play in all the games.  The boy ran to meet his sister.
  • I like this school fine.  My teacher takes a deep interest in my work and helps me whenever I need it.  I have made lots of good friends among the students and I trust I will meet them often in coming years.
  • The typographical error involves the final s in the second example: it’s shown “heavy” but, based on my reading, should be light.

Quinn Stenotypy:

  • Rich and rare were the gems she wore,
    And a bright gold ring on her wand she bore;
    But O her beauty was far beyond
    Her sparkling gems and snow white wand.
  • Nothing more fully attests the utility of shorthand writing than the rapidity with which it has grown into use during the past few years.  Fifty years ago a skilled shorthand writer was rare, and the art was looked upon as a sort of mystery, and so formidable of attainment as to deter all but the most courageous from attempting to fathom its mysteries.

*Postscript, January 13, 2019: Since posting this, I’ve learned about some comparable systems of later date, most notably the Steno-Short-Type System (1923) of Emma Dearborn, better known under its later trade name, Speedwriting, and later refined by John Sheff, Joe M. Pullis, and others.  Like some of the systems covered above, Speedwriting was developed for both typewriting and handwriting—thus overlapping with many other systems of “abbreviated longhand”—with various so-called “ABC shorthand” systems put forward in competition with it besides.  But there seems to be virtually no cross-referencing between accounts of these later systems and the ones I’ve described here.  In any case, the serious teaching of QWERTY-keyboard stenotypy plainly continued later than I’d thought.

One thought on “Shorthand on the QWERTY Keyboard, 1875-1917

  1. Pingback: All Griffonage That On Earth Doth Dwell | Griffonage-Dot-Com

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.