The Digital Revolution ?¢‚Ǩ‚Äú A series of Disruptive Events

Since the invention of the Gutenberg printing press the world has seen a continuously increasing rate of technology driven change resulting in the highly technology driven environment we live in today. It is fair to say that the Gutenberg printing press caused a revolution in printed material. It was a disruptive event. Mechanizing the printing process forever changed not only the process of printing but the availability of affordable printed material. The first book to be printed, the Bible, now accessible by the masses likely spurred on the Protestant revolution. Students at Universities also now had access to a breadth of scholarly printed material, in business contracts could be standardized, documents printed and for all the newspaper industry came into being.

A disruptive event is a change in technology that has a pervasive, long lasting, cascading and irreversible effect on the direction of technology, on society and the status quo.  Some effects are subtle e.g. the long term effects are visible but the cause may not be readily identifiable, while others have an immediate and visible impact.

This is the first of a series of planned articles discussing the impact of the digital technology revolution as a series of disruptive events that continue unabated, even accelerated, today.

While there are a number of candidates who may be identified as the creator of the first major digital event, none in my mind rises to the level of Herman Hollerith (1860-1929).  Hollerith worked at the United States Census Bureau from 1879-1882 and became engrossed in finding ways to reduce the time and labor involved in census analysis by automation.  His objective: find a solution for the 1890 census.  Why was this effort important?  The last census (1880) was not completed until 1887.  Seats in the House of Representatives are apportioned based upon population as determined by the last census so the 7 years to complete the census was totally unacceptable having left many states in limbo regarding their representative allocations.  To compound the problem, the US population had increased by more than 12 million since the last census and the shift of population westward was well underway still further increasing the difficulty of taking the census.  

To solve this dilemma the Census Bureau decided to sponsor a competition in order to find a way to dramatically reduce the time required for taking and processing the census data.  Hollerith, in his travels by rail, had observed the use of what was called a "punch photograph".  Conductors would punch notches in the edges of a passenger's ticket which physically described the passenger in sufficient detail to allow only the described passenger to use that particular ticket.  A passenger's height, hair color, build, eye color, etc. could all be coded onto such a ticket by punching the indicated places around the edges of the ticket. Hollerith theorized that a similar scheme might work for the census.  He designed an approach and was awarded two patents, one in 1884 to use holes punched in paper tape to record and mechanically read that information, and the second in 1887 using holes punched in a paper card instead of tape.  The card needed to contain enough information to be useful for the census, so he settled on using the size of a US banknote; apparently he also saw an additional benefit by fitting cards in the same size boxes as used by the Treasury Department.

Herman Hollerith won this competition handily using his system of punching, reading and sorting the cards. The official 1890 population count was completed in three months and the total population of the United States was found to be 62,622,250.  In another two years all the census data was processed and organized.  The cost was $5 million below forecast and saved more than two years time.  He had not only solved the problem of the 1890 census but his solution was the basis for compiling census data for the next 70 years.

 
Shown above is the card tabulator with 40 dials for counting, one for each of up to 40 hole positions in the card. The pantograph punch for punching cards one hole at a time is on the left side of the table, a card reader is on the right side of the table and the sorting machine stands to the right.  The dials were each wired to 40 possible holes in the card reader.  In this early implementation each possible hole represented something to be counted.  It could be a state, county, city, country of origin, family size, ethnicity, age, sex etc. The pantograph had a mask which defined the content of each hole. Every time the reader sensed a hole the appropriate counter would increment and a pocket of the sorter would open according to the category identified by the holes punched.  An experienced operator could tabulate up to 80 cards per minute and then transcribe the results from the dial counters.  The initial card used 22 columns with 10 punch positions available per column.  This was gradually increased until it reached 45 columns with 12 positions per column.  A print of a card from the 1900 census is shown below.

 
The cards were pre-printed with fields outlined according to the application.  Forty-five columns were available in this design.
 

The picture above shows more detail regarding the design of the machinery.

Hollerith incorporated his business as the Tabulating Machine Co in 1896. Interestingly; the company did not sell but rented its equipment and maintained the exclusive right to produce the cards. This did not change until the Justice Department sued IBM in the 1950's to make the tools for the manufacture of the cards available to competitors.  In 1911 the company merged with other companies to form the Computing Tabulating Recording Co.  In 1914 Thomas J. Watson joined the company and, through his leadership, dramatically enhanced its business.  In 1924 the name was changed to International Business Machines, becoming the IBM we know today.  The card design was extended to 80 columns, the holes were made rectangular and the design patented in 1928.  It remains the standard to this day.  Over time several attempts were made to either copy the Hollerith design or come out with alternatives but they never achieved the success of IBM.  Most notable of these was the Remington Rand Co. that extended the original card design to 90 columns by splitting each physical column into two groups of 6 holes each.  While Remington Rand achieved some success they were never able to match IBM in the marketplace.

What the punched card did was create a document that could be a basic business and accounting instrument regardless of the size of the organization.  A card was able to represent for example an account, a transaction, a part, an invoice, a check, a bond or a security and as a result businesses and governments could begin to automate their fundamental processes.  It eventually became known as a 'unit record' and IBM named the line of equipment designed to process the cards as Unit Record Equipment.  

The period after 1890 was one of dramatic economic growth putting great stress on businesses to keep up with the changes.  The transition to punched cards greatly facilitated the task of keeping up with sales, orders, inventory and accounting procedures.  The government used hundreds of machines during World War I to keep inventory, medical and psychological records.  So successful was this transition to punched cards that by 1918 over 100 million cards a month were being produced.  As the applications grew in volume the machines handling the cards became more and more sophisticated, ultimately being able to produce reports by the thousands of pages and process cards up to 1000 cards a minute.  All data was transcribed to punch cards. The card became the perfect medium for entering, manipulating and storing data.  There was no other medium that so perfectly encapsulated business information.  By long before the 1950's every industrialized nation was totally dependant upon these cards and the equipment that processed them.  IBM was a juggernaut in the marketplace not only because it had the lion's share of the business, but it also owned all the equipment due to its rent only policy and was the sole producer of the punched cards.  Cards had become the universal medium for handling data, and IBM machines processed the data.

This set the stage for the computer revolution that was to come.   When the computer arrived cards were the only vehicle for storing and maintaining programs.  Cards were the only source of data and until the 1970's when direct entry through terminals became available. Until then cards were the only way to feed the now prevalent computers.  More about this phase later.

In summary, Herman Hollerith was truly a giant when measured by the impact of his invention.  The introduction of the punched card was a disruptive event that changed the course of history.