Edmund Storms wrote:

That's true, but ancient economies were pretty complicated!

What standard would you use to judge? Surely, past economies were not as complicated as what we see today.

I do not know much about economics, but premodern manufacturing was, in some ways, even more complicated than most present day versions, because it required so much manual expertise and input from so many different people. A good example is the 18th century British chronometer (marine navigation clock). The complexity, quality control, the number of people involved, and cost to manufacture rivaled that of the minicomputer circa 1980.


Until modern computers came along, all recording had to be done by hand. Of course methods were developed to make simple calculations.

Not just simple ones! Remarkably complex calculations were performed manually, by coordinated groups of hundreds of clerks. Good examples are the US Census Bureau statistics from 1860, which came in four volumes with hundreds of tables, and the lunar navigation tables produced by the British Admiralty. For that matter, the first atomic bombs was developed using only human "computers" equipped with mechanical calculating machines.

An interesting point is that many of the techniques used in computer data processing evolved directly from the manual techniques used by live clerks. This is reflected in computer terminology such as "ledger," "register" "record" and "index." The ISAM index algorithms I implemented back in 1980 were derived directly from manual systems still used to shelve books in libraries.


The business models used today use equations that would take years to solve using these tools.

And for that reason, back in 1860 (and 1940) they used more elegant and clever equations and tools, that still gave reasonably good answers. My mother, for example, used a slide rule to the end of her days.


The [1880] data took 9 years to process. Since the census must be taken every 10 years, according to the Constitution, that was the limit.

Nine years to process just a few questions!!

No, lots of questions. To start with they recorded a fairly large set of basic information for every citizen: age, sex, race, national origin, address (town, county, state), trade (I think), and before 1865, legal status (free or slave -- as mandated in the Constitution). I do not think the basic census form today is much more complicated. They broke this down into six major tallies, with various actuarial of tables and so on.

Furthermore, starting in 1810 the Congress ordered the Census Bureau to collect more extensive information from some people and most commercial establishments, relating mainly to commerce. The Congress mandated "a statistical report . . . covering the kind, quantity, and value of goods manufactured, and the number of manufacturing establishments in each State, Territory, district, and county" for 25 broad categories of manufacture, and 220 kinds of goods. In 1820 this was expanded to include data on "the location of establishments, 14 additional inquiries elicited information on raw materials employed (kind, quantity, and cost), number of employees (men, women, and boys and girls), machinery (whole quantity and kind of machinery and quantity of machinery in operation), expenditures (capital, wages, and contingent expenses), and production (nature and names of articles manufactured, value, demand, and sales)."

Believe me, it only got more complicated after that!

See: http://www.census.gov/epcd/www/pdf/giq97/GIQ_history.pdf

And if I may add a blatant plug for statistics and the Census Bureau, the U.S. has an old and proud tradition of gathering and publishing accurate, detailed data about every aspect of our society, such as public health, manufacturing, employment and so on. This transparency is one of our great strengths as a society. Traditionally, only the U.K. and France have been as forthright.


The information being stored about everyone, on which taxes, employment, and credit are based, would not have been possible until recently.

On the contrary, the Constitution mandated that the data be collected for everyone, in order to divvy up the representatives in Congress, and to impose taxes and distribute benefits fairly. That is why they established the Bureau in the first place. You cannot have democracy without good, solid, detailed statistics, and the U.S. had 'em starting in 1790.


It took a lot of manual work to do an in-depth analysis of the data sets, for things like actuarial tables, but 19th-century statisticians did impressive work.

No doubt they were impressive given the tools available. What would you expect to happen to society if the computer had not been discovered?

I know exactly what would have happened. The cost of the 1890 and 1900 census would have gone through the roof, and the Congress would have demanded someone come up with a mechanical solution. That is exactly why and how the computer was invented, by Herman Hollerith. There were several competing designs but his was best. His company evolved the IBM corporation.

When a society feels a strong need for a tool, and the tool is technically within its grasp (meaning it does not require any fundamental new discovery), development becomes inevitable. The ancient Egyptians felt they must have pyramids, so they devised ways to build them. The British in 1750 decided they must have accurate chronometers, so they created a high tech industry to supply them. In 1890, the U.S. Congress understood that the survival of democracy depended upon mechanical data processing, and it decreed that it would pay for this technology, so inevitably someone invented it. That does not mean the Congress could today decree that someone figure out the secret of cold fusion. The Hollerith machines were clever, elegant and well designed, but they did not require any fundamental breakthroughs in physics, mathematics or electricity. They were based on existing technology and well-understood principles.

On this issue, Arthur Clarke went even farther out on a limb than I do. He wrote: "Anything that is theoretically possible will be achieved in practice, no matter what the technical difficulties, if it is desired greatly enough."

- Jed


Reply via email to