1

Ch. 1, Inventing the Pretty Typewriter

Virginia Woolf was fascinated by the gap between the real and the fictional, the world and the word. They were distinct, yet never wholly dissevered. “Fictionis like a spider’s web,” she wrote, “attached ever so lightly perhaps, but still attached to life at all four corners . . . attached to grossly material things, like health and money and the houses we live in.”1 If we are to probe the fictional webs that came to be spun around the modern secretary, the hundreds of novels and films in which she is the heroine, we must attend to that world of “grossly material things” in which those stories took shape and made sense.

1

One “grossly material thing” that connected millions of typists, secretaries, and stenographers was plainly the typewriter. Traditional histories credit its invention to a single man, Christopher Latham Sholes (1819-1890) (fig. 1), at one time the editor of a small-town newspaper in Kenosha, Wisconsin,and a sometime dabbler in politics. In his editorialshe took positions that were idealistic, downright impractical, or forthrightly progressive: he urged abolition of the death penalty, demanded the elimination of war, and staunchly supported equal rights for women. In 1848 he was elected a state senator, then served a brief term as city clerk of Kenosha, and in 1851 returned to the state legislature as an assemblyman. In January, 1853, he met James Densmore (1820-1889) (fig. 2), then the editor of the True Democrat(a newspaper in Oshkosh, Wisconsin), a meeting that would ultimately have a profound effect on his life. Months later the two men attempted to transform Sholes’s weekly Telegraph into the Daily Telegraph,a venture that ended in failure a year later when the price of their wire service from the Associated Press became too high. The two parted amicably: Sholes stayed on to edit his weekly, slightly renamed Tribune and Telegraph in Kenosha while Densmore moved to a nearby town and another newspaper. But it was not the end of their story.

In 1857 Sholes moved to Milwaukee, where he worked as a journalist first for the Free Democrat and later The Sentinel, both Republican newspapers. When the Civil War broke out, Sholes volunteered, at his own expense, to be the governor’s personal representative and report on the physical care of Wisconsin soldiers serving in the Union army. As a reward he was named collector of the port of Milwaukee in 1863, a sinecure that entailed only light duties and consumed little time. It enabled him to quit the newspaper business and devote himself to his growing interest in inventingmachinery, a penchant he had indulged ever since his move to Milwaukee. In 1860, working with Samuel W. Soule, a draftsman and civil engineer, he had designed a machine for addressing newspapers sent to customers subscribing by mail, one that was soon successfully manufactured. He had also designed another machine for numbering documents requiring successive pagination or enumeration, such as ledgers, tickets and coupons. He patented it in 1864, together with improvements made in 1866 and 1867. It was at this point that he came across an article in the journal Scientific American of 6 July, 1867, which reported on a “Type Writing Machine” that had recently been invented. It prompted Sholes to wonder whether he might be able to devise a better one.2

1

Working once again with Samuel Soule and also with a lawyer named Carlos Glidden, Sholes set out to build a working model of a machine that would embody his essential insight. The writing machine described in Scientific Americanhad characters that were arranged on a wheel that would rotate a character into place, supplemented by a hammer that then stuck the paper against it. The procedure entailed so many steps to produce a single letter that it could never be faster than normal writing. Sholes, instead, decided that he would put each type or character on a separate bar, and that each bar would individually strike the paper: a single motion for a single character. By September, 1867, he had a working model and even arranged for a demonstration. Charles E. Weller, chief operator at the local office of Western Union telegraph, was so impressed that he put in an order for the very first one to leave the shop. He would receive it only four months later: Sholes had devised the working model of a potentially useful machine; but reproducing it by manual production was slow and costly.

To remedy this problem, Sholes and his colleagues decided they needed capital and manufacturing expertise. Soule was sent to New York and Washington, lugging a bulky model with him, but failed to find a backer. Sholes, instead, recalled his former colleague and friend James Densmore; he sent him a letter, typed with the new machine, in which he described the invention and hinted at its possibilities with a quotation from Shakespeare:

There is a tide in the affairs of men

Which, taken at the flood, leads on to fortune.

1

Densmore, now living in Meadville, Pennsylvania, and working as an attorney for amachine company, was immediately interested. Weeks of negotiation by correspondence led to an agreement in November, 1867: in return for paying Sholes, Soule, and Glidden $200 each, Densmore would receive a 25% share in the venture and undertake to finance the machine’s manufacture. In the years since his collaboration with Sholes in Kenosha, Densmore had also become an inventor, having devised and patented an early version of a tank car for transporting oil on railroads. His set of skills was suited to the typewriter enterprise: former editor and publicist, inventor, and lawyer. In March, 1868, he finally arrived in Milwaukee,took command of the project, and swiftly discovered that he was part-owner of a machine much cruder than he had thought.He immediately demanded improvements.

In late June that year he journeyed to Washington, D.C., and filed claims for two patents:one covered the design of the earlier machine demonstrated back in September, 1867; the other, the improved machine developed over the last three months. Densmore now took the newer, improved machine to Chicago. In association with Soule and E. Payson Porter (who ran a telegraph school), he spent $1,000 to manufacture fifteen machines. Alas, when tested at Porter’s school, they swiftly jammed. Densmore abandoned plans for further manufacturing and instead returned to prodding Sholes and his colleagues to make more improvements.

The next stage of tinkering required more than two years, and it was only in March, 1871, that Densmore felt he could undertake a second attempt at manufacturing. This time he used a machine shop in Milwaukee owned by Charles F. Kleinsteuber (fig. 3); it included a brass foundry as well as general machine tools for model making. The newly manufactured machines worked well enough at the beginning, but after intense usage some of their characters fell out of alignment. Production was halted after making only twenty-five machines, and once more Densmore directed his team of inventors to find a solution.

1

By the spring of 1872, Densmore was again convinced that the machine was technically ready; he also decided that he would personally supervise a third attempt at manufacturing, this one to take place in Milwaukee but at a different location, an old mill with water power. By late June he had installed second-hand tools, chosen a superintendent, and hired several laborers. But the machines were still being handmade. Densmore inspected each one and often required alterations or the remaking of parts. Even he could see that the scale of the Milwaukee operations was inappropriate: “I am anxious to get everything in such shape that the various parts can be made by machinery, without this everlasting filing and fitting, which makes but a botch after it is done.”3 The machines were a technical success and a financial failure. “The fact is that as we are now making them they are costing more than we ask for them. And until we cheapen the making, we are losing all the time,” he complained.4 It was in November, 1872, that Densmore and Sholes decided to make one last change to the machine, altering its keyboard in a way that would make it less likely to jam by separating the most frequently used keys, resulting in the famous QWERTY configuration--so named after the first six letters in the topmost row of keys--that hasbecome the so-called “universal” keyboard still in use today. Every time that we use or office PC or home laptop, our fingers retrace the steps that were taken to solve a problem that hasn’t existed for more than thirty years: preventing the typewriter bars from jamming as the rose toward the central point where the keys struck the page.

One month later, in December, Densmore was visitedby a friend who bore the sonorous name of George Washington Newton Yost; it wasYost who suggested to him that he take the typewriter elsewhere for manufacturing. He recommended the arms makers Eliphalet Remington & Sons of Ilion, New York, and offered to accompany him there.Some two months later, in mid-February, 1873, Densmore and Yost journeyed to Ilion.A picturesque town with a small creek that flowed into the Mohawk River, the largest tributary of the Hudson,it was dominated by the massive Remington factory, so vast it would cover five acres if spread out on a single floor (fig. 4). Staying at a small hotel, Yost and Densmore were greeted by three men: Philo Remington, the eldest of the founder’s three sons and the firm’s president since their father had died in 1861; Henry Benedict, a young executive; and Jefferson Clough, the superintendent and head mechanic of the Remington works. Yost and Densmore gave the men a working demonstration of their machine. Fifty years later, Benedict recalled:

We examined and discussed the machine for perhaps an hour-and-a-half or two hours and then adjourned for lunch or dinner. As we left the room, Mr. Remington said to me, “What do you think of it?”

I replied, “That machine is very crude, but there is an idea there that will revolutionize business.”

Mr. Remington asked, “Do you think we ought to take it up?”

I said, “We must on no account let it get away. It isn’t necessary to tell these people that we are crazy over the invention, but I’m afraid I am pretty nearly so.”5

One part of Benedict’s recollection may have been affected by the half-century of events that had subsequently intervened, his confident assertion that the typewriter would “revolutionize business.” If that is what he truly said to Remington, he was extraordinarily prescient. Sholes and Densmore, instead, thought it useful for telegraph operators wishing to convert Morse code into legible text, which is why the first demonstration of the machine in 1867 had included the telegraph operator who then ordered the first machine to be produced, or why Densmore’s first manufacturing attempt in Chicago had been in collaboration with a man who ran a school for training telegraph operators.6 Who would use the typewriter was still an unanswered question that lingered in the background during the two weeks of negotiationthat followed. On 1 March, a contract was at last signed, with terms that greatly favored the Remingtons. Densmore would pay them $10,000 in advance, and also grant a royalty of $0.50 per machine to Jefferson Clough, the Remington head mechanic who would redesign the machine for mass manufacture. The Remingtons agreed to produce at least 1,000 units, plus a further 24,000 at their discretion, and would receive a fixed price for each machine manufactured (the exact sum is not known).7

Densmore now had to find cash to meet these new obligations. He borrowed $3,000 from Clough, enough to keep a few creditors at bay, then returned to Chicago to hustle up $10,000 from Anson Stager, a telegraph entrepreneur who, in exchange, received exclusive rights for selling typewriters in certain western states for the Western Electric Company. It was yet another financial complication--just when Densmore thought he had solved them.

1

In November, 1872, four months before Densmore signed the agreement with the Remingtons, he had added up his own investment in the typewriter and found that he had spent $13,000 in financing experiments, patents, and manufacturing.8 Most of it was money he had borrowed against the future royalties that would accrue to his 25% share of the project, including substantial loans advanced to him by his brothers Amos and Emmet. It was time to reach a comprehensive settlement that would not only convert these loans from liabilities into assets, albeit passive ones, but also formalize other transactions that had altered the financial relations among the four co-owners of the patent claim: Christopher Latham Sholes, Densmore himself, the draftsman Samuel Soule, and the lawyer Carlos Glidden. Soule, during the intervening years, had moved to New York and agreed to sell Densmore his 25% share for $500 (which Densmore duly purchased, though his actual payments were extended over time). Glidden, instead, had drifted away from the typewriter project, absorbed in another invention, and to finance work on it had offered Sholes his 25% share inexchange for his serving as co-signer for a note of $250. When Sholes hesitated, Densmore instructed him to go ahead and promised that he, Densmore, would pay for the note, as he later did. Densmore, in short, owned 50% of the original patents and co-owned 25% with Sholes; while Sholes owned his own 25%. Glidden, meanwhile, had returned to the typewriter project and helped Sholes in further developing it between 1870 and 1872. He was therefore justified, or so he thought, in claiming back a portion of the share he had sold. Densmore rejected this claim; he himself had paid Glidden a salary for working with Sholes, and in his view owed him nothing further. Glidden then turned to Sholes, asking that he give him a portion from his share in the business. Sholes had so little faith in the commercial viability of the typewriter that he was happy to placate Glidden with a portion from his own share.

The formalization of these arrangements was an “agreement of trust” that Densmore drew up and executed on 16 November, 1872. The various owners of all interests in patents already obtained, pending, or to be applied for, agreed to assign them to Densmore and Sholes jointly, as trustees. The trustees could make and sell instruments or license others to do so, and all proceeds would be divided as follows:

James Densmore40%

Christopher Latham Sholes30%

Carlos Glidden10%

Amos Densmore10%

Emmett Densmore10%

1

But in April, 1873, less than a month after Densmore had contracted with Remington, Sholes decided to sell off one of his three remaining 10% shares;he mistrusted the typewriter’s commercial prospects, and his present needs were pressing. He disposed of it in fractions to Amos Densmore and two others for a total of $5,350, partly paid in cash. Densmore grew alarmed: outsiders buying shares from Sholes might create trouble for his ongoing management.In response he formulated a new plan: together with his brother Amos and G. W. N. Yost, he would buy out Sholes and any other shareholders. To do so he created a new firm, Densmore, Yost & Company, one that promptly repurchased all the portions Sholes had sold to outsiders, as well as Emmett Densmore’s entire tenth, and one of Shole’s two remaining tenths, the latter bought with $10,000 in promissory notes. The agreementwas struck in September, 1874.Sholes was now left with only a single share, ten percent of the whole.9

Meanwhile, on 30 April, 1874, the first Remington-made machine finally arrived at the sales office that Densmore had opened in New York. It was a handsome piece, enclosed in metal painted with glossy black enamel, and distinctly resembling a sewing machine, replete with a foot treadle (such as contemporary sewing machines had) that shiftedthe paper from one line to the next (fig. 5).Two months later six more machines were shipped to Washington and ten to Chicago, all for use by shorthand reporters who would pay for themby writingglowing testimonials. Things were up and running—at last. Remington announced it was ready to produce as many as a hundred machines in the next month. But itsestimate of production capacity bore no relationship to real demand. By the end of 1874 only 400 machines had been sold. And figures for the next four years were hardly better, averaging 900 per year. Between 1874 and 1878, four thousand typewriters were manufactured and sold.10

1

Tellingly, that figure broadly coincides with another figure from the U.S. census for the year 1880, one that shows only 5,000 people employed as stenographers or typists in the United States. But it also shows something else more startling: already 2,000 of them (or 40%)were female. Those figures are still more startling when compared to the corresponding figures from the census of a decade earlier in 1870, when the same occupational category had contained only 154 workers in total, of which a mere 7 were female (or 4.5%). Not only had the number of total workers increased by more than 3000% in the intervening decade, but the percentage of female workers had multiplied by tenfold.11 It signalled the beginning of a revolution. And yet it was also a revolution that a contemporary might have failed to discern. For while the census of 1880 was plainly conducted in the year 1880, the labor of compiling and collating its results was so vast that it was not completed and published until ten years later, in 1890. The delay was a sign of a broader development: private enterprise was growing so large that government efforts to monitor it were hamstrung by inadequate data retrieval and collation technologies. The typewriter would become a critical instrument for more rapid and efficient data production in large corporations; but this use was apparent to neither Sholes nor Densmore. Densmore had poured his energies in getting the typewriter manufactured; but his ability to bring it to market was hampered by his inability to imagine it being used by a commercial sector, rather than isolated individuals. The machine, early advertisements urged, was suited to lawyers, clergymen, editors, and court reporters. But those made up a tiny market when compared to the business sector that would soon adopt it. Yet already in 1875 there was a telling hint of the future. The firm of Dun, Barlow & Co. (predecessor of the celebrated business information firm, Dun & Bradstreet) purchased one hundred typewriters to equip its main office, then sent another forty to its branches throughout the United States, together with instructions that all reports now had to be typed. Previously subscribers who sought out the firm’s business evaluations had to go to its offices to consult a handwritten ledger; now typewritten reports would be routinely mailed to them.12 It was data intensive firms such as Dun Barlow that constituted the real market for the typewriter.