Of typewriters, runes, and short-hand

Or, why is the & on the 7 key?

Back in ca. 2003, when I was preparing the diplomatic edition of Beowulf for Beowulf on Steorarume (heorot.dk), I noticed the use of what looks like the Arabic numeral ‘7’ for and (Old English ond).

Beowulf f.132v: ll.120b-123a

    wiht unhælo
    grim grædig gearo sona wæs
    reoc reþe on ræste genam
    þritig þegna

    The unholy creature,
    grim and greedy, soon was geared-up,
    wild and savage; and seized from their sleep
    thirty thanes

This turned out to be a Tironian et (⁊) , one of a number of shorthand symbols devised by Marcus Tullius Cicero’s slave and ppersonal secretary Tiro in the first century which continued to be used by monastic scribes in the medieval period. The more familiar abbreviation of et “and” is of course the ampersand &, which, unlike , is in fact a stylised et. (For more discussion of some of the back history of both the ampersand and the Tironian et see these Shady Characters posts.)

Other shapes of the Tironian et

The fact that the Tironian et , a shorthand ‘quick’ symbol like the ampersand &, looks like an Arabic numeral 7 makes the appearance of the & above the 7 (i.e. as the shifted value of the 7) seem significant. At the time of creating the diplomatic web edition of Beowulf I assumed that this must be the reason for the particular placement of the & on keyboards.

Unsurprisingly, other people have thought of this as well, including a StackExchange question (as well as a comment on a Shady Characters post on ampersands). There the ‘accepted answer’ to this question says:

Although the 7 was the ampersand on IBM’s standard keyboard layout, that is hardly universal. The first nine printable characters in ASCII are ! " # $ % & ' ( ), which should give a good clue as to what the top row of a teletype keyboard looked like. On many early teletypes and terminals (and also, BTW, on the Apple ][), the shift key toggled bit 4 of the character being produced, thus it would turn a 1 (011 0001) into ! (010 0001), and , (010 1100) into < (011 1100). Since the digits 1–9 received consecutive code, so did the characters produced by combining them with the shift key. Shift-7 on those keyboards was apostrophe; the ampersand was shift-6.

Other typewriter keyboards also varied considerably in where they put the ampersand. Its association with the number 7 is nowhere near as consistent as the association between 1 and !, 3 and #, 4 and $, or 5 and % which existed in both older computer keyboards and today’s US arrangement.

The association of ! (‘bang’) with 1 actually has a fairly tidy explanation as many of the early typewriters omitted the numeral 1 and the typist was expected just to use the lowercase letter l in its place; likewise the ! was omitted and was also typed using the lowercase l then backspacing and typing a full-stop ..

While ASCII and older computer keyboards may not provide evidence for early association of 7 and &, an examination of the early typewriters from antiquetypewriters.com reveals firstly that almost all of the early typewriters collected there which have symbols as the result of ‘shifted’ numerals do in fact place & with the 7 including the 1892 Norths typewriter (London), Wagner Typewriter Co’s (New Jersey) 1896 Underwood 1, the 1898 Shimmer (Milton, Pennsylvania), Meiselbach Typewriter Co.’s (Kenosha, Wisconsin, USA) 1901 Sholes Visible, and, secondly, and crucially, the 1878 Remington Standard Model No. 2 (New York) also positions & as the shifted value of 7. The Remington Standard No. 2 is described as the first commercially successful typewriter, and its (QWERTY-style) layout thus became established as standard.

Remington Standard No. 2

The one early typewriter I found with shifted numerals as symbols which didn’t associate 7 and & is the 1895 Waverly (London) where the & is above the 6 and the 7 has a / above it.

So an association between & and 7 for keyboard does seem to have been established in the 19th century, presumably through the success of the Remington Standard No. 2. The question then shifts to whether or not the placement of the & above the 7 by the designers of the Remington Standard No. 2 was motivated by Tironian sensibilities or not.

The Tironian et has disappeared from most everywhere, excepting in modern Irish typography. Interestingly, some Irish typewriters place the Tironian et above the 7 key.

Keith Houston at the Shady Characters website suggests that:

the Tironian notes suffered near-extinction in the Middle Ages, victim of a curious linguistic witch hunt. The secrecy and cipher-like nature of both traditional runic writing and short­hand did not sit well with the distrust of witch­craft and magic prevalent in those times, and Tiro’s system fell out of use.

Indeed, in Old English manuscripts, runes sometimes appear, standing in for the words which are borne by the particular runes as names, including (mann “man”), (dæg “day”), and (œþel “ancestral land”). The last of these is used three times in the extant text of Beowulf.

Beowulf f.141v: ll.520–521

    ðonon he gesohte swæsne ·ᛟ·
    leof his leodum lond brondinga

    then he sought his own sweet ·ancestral homeland·
    – loved by his people – the land of Brondings

In any event, what of the question of “why is the & on the 7 key?”

It would seem in all likelihood that the placement of the & on the 7 is a happy accident. Outside of Ireland, Tironian ⁊s would have been long enough out of fashion that it seems unlikely anyone (or, at least, typewriter engineers) would have associated & with 7. The positioning of the Tironian above the 7 key on some Irish typewriters, on the other hand, is then doubly predestined, given the largely standard positioning of the semantically-equivalent & over the 7 on many typewriters and the similarity in appearance of the Arabic 7 numeral and the Tironian symbol.

Benjamin Slade
Benjamin Slade
Associate Professor of Linguistics

My research interests include formal semantics and syntax, historical linguistics, South Asian and Caribbean languages, and the use of computational concepts in formal linguistics.