[Journal of the Simplified Spelling Society, 1994-1 pp24-28 later designated J16]
[See Bulletin Fall 1982 Shorthand.]

The Palantype System: another readable shorthand of the English language.

Patricia Thomas.

Patricia Thomas is a free-lance terminologist and editor of biological journals, currently doing research for her PhD on special language terminology. She learned to Palantype in English, French and German at the college of the Lycée Français de Londres and has used the system at conferences and in multilingual office situations.

PALANTYPE AND MACHINE SHORTHAND.

The review on AgiliWriting (AW) by Chris Upward (J14 1993/1, pp29-33) explains how an abbreviated, alphabetic shorthand can, if taken down correctly and legibly, be transcribed by anyone else on a computer which has an automatic word processing conversion program to transcribe the text into standard English. The method is reminiscent of a machine shorthand system, Palantype, which has been in use for many years, and by which a stenographer - in this case, a Palantypist - takes down speech directly on to a special keyboard; hence the letters start their life in a printed form. The system has in recent years been radically developed with the result that Palantype too can be automatically transcribed into standard English via a word processing package. The Palantype method initially appears to have the advantage of accuracy because the initial input is always in printed form and there is therefore less risk of error than from hastily handwritten abbreviations. However, the advent of the lap-top computer and its use by Agilitypists renders this problem obsolete (Upward, 1993/1, p30). An initial disadvantage of Palantype compared with AW was at the basic level, because the cost of AW can be very low; there can surely be no method cheaper than a pencil and paper! And because it can be handwritten, it is a boon to people who at times work outside, such as journalists. However, now that AW is supported by computer hardware and software development, both systems have become more expensive to operate.
A pre-electronic Palantype machine in action.

Fig.1: a pre-electronic Palantype machine in action.



An aid for the hard of hearing.


One very important advantage of Palantype is its enormous benefit to deaf people, who were one of the early targets for the renewal of interest in the system. (A trained palantypist sits beside the deaf person who reads the spoken text on a screen almost immediately after the speaker has spoken.) However, this is perhaps not a just comparison to make because the developers of AW may not have wished to explore this avenue.

Accuracy.


The main aim of AW and its later forms, AW+ and Agilityping, is to produce a shorthand readable by anyone trained in the method, primarily for office use. The improvements of AW+ over AW are that no word form exceeds five characters and each homophone or near-homophone has a different Agiliform, hence "advice" = ADVS and "advise" = ADVZ; "their" = THYR and "there" = THR. Each AW word form is matched against a dictionary of c.60,000 word forms, with claims of 100% accuracy (Anne Gresham, 1994, personal communication). The system runs on both IBM and Apple Macintosh.

Palantype input can be simultaneously transcribed into English, the spelling checked against a stored lexicon which was originally in excess of 71,000 inflected English word forms (although most Palantypists now find dictionaries of c.20,000 words sufficient for their needs), and the result displayed on a large screen or VDU. Hard copy can be printed concurrently and the input stored on an IBM-compatible disk (the system runs on PC-DOS or MS-DOS version 3.3 or later). Accuracy against the dictionary match is 90%, the remainder responding to grammar and word structure rules, resulting in 95% accuracy or greater, while the rest is easily comprehensible.

Speed.


The speed of Agilityping is claimed to be around 80-120 words per minute (w.p.m.). This is much the same speed as that reached by most 'conventional' shorthand writers and may be sufficient for most offices but would almost certainly not be high enough for conference and court proceedings. This is where Palantype really scores because speeds of 180-200 w.p.m. are routinely recorded, and the system has been in use by court reporting companies for a number of years, and by the Police when recording interviews.

HISTORY OF MACHINE SHORTHAND.

Readers will no doubt be familiar with the sight, particularly in courtroom scenes shown in American films, of a stenographer sitting in the centre of the courtroom in front of the Bench, with a compact black box on her (occasionally his) lap (Fig.1). A little about the interesting history of machine shorthand, and in particular the Palantype system, will help its versatility to be appreciated.

The first shorthand machine was invented in France by Gonod in 1827, predating typing by 40 years, but it was not until 1910 that the French Grandjean machine was patented. The early systems used dots and dashes similar to Morse code, and this was true of the first British machine, the Stenotyper, used in London around 1900. The first English language version was patented at about the same time in America and called the Stenograph machine, where the one on which today's machines are modelled is that of W.S. Ireland, who developed a 22-key keyboard in 1914.

Development of Palantype 2.


Britain, in contrast, used Isaac Pitman's shorthand from 1837 and various forms of this and many others dominated the scene in offices. Machine shorthand was introduced into Britain between the two World Wars by a French teacher and psychologist, Mademoiselle Palanque (from whose name Palantype is derived). The system is based on that of Grandjean and uses mainly Roman alphabetic characters, with one or two additional symbols in the early models. It is thus easier and quicker to learn and to decipher than the purely symbolic form adopted by most shorthands. The general principle of the system is that speech is recorded phonetically syllable by syllable (rather than phoneme by phoneme or letter by letter), giving a highly accurate recording of the language. Ergonomic factors were taken into consideration and the keyboard, which has 29 keys, is designed so that the stenotypist sits in a comfortable position with the hands placed naturally on the keys. This contrasts with the position of sitting at a piano or at most computer keyboards where the hands, when close together, turn outwards slightly from the wrists. Comfort is of course important, in view of current concern about Repetitive Strain Injury.

the Palan 2000 keyboard.

Fig.2: the Palan 2000 keyboard.

Recent developments.


Research into Computer-Aided Transcription (CAT) of the Palantype system was undertaken by W. L. Price at the National Physical Laboratory (NPL) in 1967 to provide high quality recording of court proceedings. He modified the keyboard slightly to make the output more readily assimilable by the computer. Due to high computing costs, the work was not exploited commercially at the time. However, further research was undertaken in the 1970s at Southampton University, Leicester Polytechnic and the BBC. An undergraduate project was begun in 1974 by Dr. A. F. Newell and his team, which included Dr. A. Downton and Dr. C. Brooks, at the University of Southampton. The research was supported by the NRDC with the aim of producing a portable system which could be used to provide a simultaneous visual transcript of lectures and meetings as an aid for the post-lingually deaf and hard-of-hearing (Brooks, 1985: 13). A number of experimental systems were used by, among others, the British politician Jack Ashley.

In 1983, Possum Controls Limited licensed the results of the research at Southampton and developed the prototypes into the current advanced CAT systems with, among other improvements, an electronic keyboard (Fig.2). The company has also introduced a more computer-compatible method, known as the PCL (for Possum Controls Limited).

Printout from a pre-electronic Palantype machine.

Fig.3: printout from a pre-electronic
Palantype machine with syllabic spelling

HOW PALANTYPE WORKS.

Syllabic phonetic structure.


The most important principle of the Palantype system which enables high speeds of dictation to be realized is its method of recording words phonetically syllable by syllable instead of letter by letter. To effect this, several keys can be depressed at the same time, as with a chord on the piano, instead of having to be depressed one at a time in sequence. The overall saving in the number of keystrokes is shown in the Tables and the corresponding saving in time is not difficult to imagine.

The keyboard.


The 29 keys are displayed in three sets, grouped in the manner of the majority of phonetic syllables in English, i.e. consonant-vowel-consonant. The operator mentally divides words into phonetic syllables, aiming to begin each syllable with a phonetic consonant where possible, including vowel/consonant doublets such as the initial 'y' sound in 'Europe' and 'use'. The early form of Palantype was printed on a paper roll which moved forward automatically after each 'chord' was depressed (Fig.3).

This stage is retained in the current system, the roll or 'band' appearing as justified text on paper behind the screen. This is the version which is usually read back verbatim, as in court cases.

The keyboard prints the letters and symbols in an unvarying left-to-right order, as determined by the phonology of English syllable structure. The letters can only appear in the following sequence, ie syllable-initial S can only precede P, not follow it in any givn sylabl:

SC(K)PTH(D)+MFRNLYOEAUI^NLCMFRPT+(D)SH

The order is the present version as adapted by the National Physical Laboratory from the earlier version (whose different letters are shown in brackets), and as further modified by Possum. The rigidity of this order can on occasions be mildly frustrating, although the operator soon learns to break the word to form a second syllable; for example, it might seem appropriate to write under with one keystroke, UNDR, but this is not possible because the R precedes the /d/-sound (T+) on the right-hand side of the keyboard. Some words can be split in more than one way and here a knowledge and understanding of etymology is useful, because it usually provides the most logical and practical way of breaking a word into syllables, as well as facilitating the 'transliteration' process into traditional English orthography. The current keyboard layout is shown in Figure 4. It will be noted that the long front bars are divided into two sections. These are functionally identical but it is easier to have two smaller keys for electronic purposes, resulting in quieter use.

Keyboard layout, PCL Palantype System

Figure 4. Keyboard layout, PCL Palantype System

As English has more than forty phonemes and not all the letters of the alphabet are represented on the Palantype keyboard, a number of conversion principles have been adopted to provide codified forms to fill the phonetic gaps. The following four points are taken from Downton (1982: 19, 21):

1. The '+' key is used to show a voiceless letter is to be read as its voiced equivalent.
For example, P+ = B, C+ = G, T+ = D.

2. Single vowel letters are used to represent the short vowel sounds, as in mat, met, hit, hot, hut.

3. The point, signified by an asterisk '*' in this paper, is used in conjunction with a vowel to lengthen the vowel sound. Thus fit and feet are Palantyped respectively as FIT and FI*T. (The asterisk is represented on the new POSSUM keyboard as ∧.

4. The consonant C by itself always represents the hard value as in cat. The soft value in cell is represented by S.

N+ stands for 'ng' as in sing and +F is V, while MF is /w/ at the beginning of a word and F+ at the end. Additional vowel sounds are represented as follows (Downton, 1982: 20):

OU for the sound of 'oo' in moot, brute
OE for the sound 'oh' in soul, mote
OI as in soil, toy
AI as in might, height
AU as in house, now.

The letter Y in Palantype can represent either J or Y as the first consonant or vowel of a word.

COMPARISONS WITH CUT SPELLING AND AW.

One of the points of common ground with all three systems is to make redundant those letters which are unstressed or silent. It seemed interesting to make a direct comparison, using as a criterion the speed of recording discourse verbatim, as shown in the following table which takes many of the examples from Chris Upward's review article (J14 1993/1, pp30-31). The number of keyboard depressions of AW and Palantype are shown in curved brackets for comparison. The "/" sign in Palantype PCL in the following table distinguishes left- from right-hand depressions of the letters and symbols.


Traditional
Orthography
Abbreviated form (No. of keyboard
depressions in brackets)
(TO) AWPalantype PCL

CS Rule 1 (letters irrelevant to pronunciation)
breath
debt
evolve
ignore
money
perhaps
you
write
breth (5)
det (3)
evolv (5)
ignor (5)
mony (4)
praps (5)
u
wrt (3)
P+RE/TH (1)
T+E/T (1)
E/+FO/LF (2)
/IC+NO/R (2)
M/U N/I (2)
PR/APS (1)
E/U (1)
R/AIT (1)
 
CS Rule 2a (unstressed vowels before <l, m, n, r>)
abundant
bundle
doctor
under
upward
cultural
abundnt (?7)
bundl (5)
doctr (5)
undr (4)
upwrd (5)
cultrl (6)
/A P+/UN T+/NT (3)
P+/UN T+/L (2)
T+O/C ST/R (2)
/UN T+/R (2)
/UP MF/RT+ (2)
C/UL TR/L (2)
 
CS Rule 2b (vowels in certain suffixes)
-ed (past participle)
-ing
singing (neither G
pronounced as such)
-d (1)
-g (1)
sgg (3)
/T+ (1)
/IN+ (1)
S/IN+ /IN+ (2)
 
CS Rule 3 (doubled consonants simplified)
clock
answer
battle
cloc (4)
ansr (4)
batl (4)
CLO/C (1)
/A∧N (or /AN) S/R (2)
P+/A TL/ (2)


Table 1. AW spelling and Palantype keyboard depressions

Special uses of U.


The vowel/consonant doublets W, Y have already been commented on, and in Palantype these would be recorded strictly phonetically. However, U might produce some homophonic anomalies; for example, would a native of Norfolk confuse muse (ME/US) and moose (MO/US)? (The /s, z/ sounds were both represented by S in Palantype, although I understand that /+S can now denote /z/.) Other examples are pull = PO/UL and pool = PO/U∧L; stewed prunes would be STE/UT+ PRO/UNS. Similarly, with near homophones, presumably AW would differentiate between ruin (ROU/IN) and rune (ROUN), since 100% accuracy is claimed. Of course, a commonsense appraisal of the context in which the word appears should clarify any semantic problems. A second table shows some further examples of special uses:


Traditional
Orthography
Abbreviated form (No. of keyboard
depressions in brackets)
(TO) AWPalantype PCL

CS Rule 1 (letters irrelevant to pronunciation)
Europe
European
?yrp (3)
?yrpyn (5)
E/URP (1) or E/U RO/P (2)
E/UR PE∧/AN (2)
(PIAN is not possible in Palantype because I is the last
vowel to be printed, although its position is in the centre
of the keyboard and it appears to lie before A and U.)
jovial
medium
onion
union
fluent
jovyl (5)
medym (5)
onyn (4)
? wnyn (4)
flwnt (5)
+JO/E +FY/AL (2)
M/I∧ T+Y/UM (2)
/U NY/UN (2)
E/U NY/UN (2)
FLO/U∧NT (1)


Table 2. Representations of the U sound.

TRAINING AND THE FUTURE.

Manchester College of Arts and Technology, in conjunction with Possum Controls Limited (Palantype Division), offers a diploma training course in verbatim reporting for which a minimum of 5 GCSEs at Grade C or above, including English language, is required. Distance-learning courses are also available by hiring a complete package from Possum Controls Limited. It is hoped that versions in other European languages will be developed shortly.

CONCLUSIONS.

It appears that in comparing the two systems, AW and Palantype, we are looking at systems which have two different objectives, but which nevertheless have a considerable amount of overlap in their application. Palantype is undeniably faster and is therefore ideal for reporting the proceedings of meetings. It is more sophisticated and has had an important side development for deaf people. Furthermore, its applicability to other European languages makes it a particularly valuable resource in our current multilingual environment. AW, although slower, is ideal for use 'in the field'. It is not known whether it could be adapted to multilingual use, but similar systems may have been devised for other languages.

Footnotes.

[1] I owe most of the historical detail to Palantype, a division of Possum Controls Limited, from material kindly provided by Dr Colin Brooks, Possum's Associate Research Director.

References.

Brooks, C P (1985) Computer Transcription of Written Shorthand for the Deaf, PhD Thesis, University of Southampton.

Downton, A C (1982) Simultaneous Transcription of Machine Shorthand for the Deaf, PhD Thesis, University of Southampton.

Upward, C (1993) review of 'AgiliWriting - the readable shorthand of the English language' in Journal of the Simplified Spelling Society, 1993/1, pp29-33.


Fig.1 - Working position of early Palantype machine: acknowledgement to Catherine Benjamin and Photographic Services at Aston University, Birmingham, for the photograph.

Fig.2 - Current Palan 2000, reproduced by permission of Possum Controls Ltd.

Fig.3 - Early printed Palantype format, from Brooks (1985), p12.

Fig.4 - Current Palantype keyboard layout, reproduced by permission of Brooks (1985: 10).


Back to the top.