6
$\begingroup$

Suppose that a mysterious philanthropist provided fundings for Charles Babbage's intricate projects and the Analytical Engine was finished by 1850.

Surely there would be a lot of raptured academicians and some fuss in newspapers, but what are some problems from 19th century the Engine could solve and thus provide substantual economic boost?

$\endgroup$
4
  • 2
    $\begingroup$ This seems to be a very broad question, effectively an open ended invitation for people to brainstorm anything that could be solved by a Babbage engine. As a rule questions asking for brainstorming, or idea generation, or questions that have many valid answers are not a good fit for this site and likely to be closed as too broad. $\endgroup$
    – sphennings
    May 24, 2022 at 13:16
  • 1
    $\begingroup$ What did you have in mind specifically as you're narrowing the question? It's also slightly tricky from the point of view that the engine wasn't all that capable compared with trained mathematicians. It was a great proof of concept, but would have been too expensive (and big) for everyone to have had one in their pocket. Needed specialist training to program as well. $\endgroup$ May 24, 2022 at 13:21
  • $\begingroup$ @JiminyCricket. That sounds a lot like a description of Eniac... $\endgroup$
    – Zeiss Ikon
    May 24, 2022 at 13:38
  • 4
    $\begingroup$ This is the premise of The Difference Engine by William Gibson and Bruce Sterling. $\endgroup$ May 24, 2022 at 15:12

5 Answers 5

11
$\begingroup$

"What are some problems from 19th century the Engine could solve and thus provide substantual economic boost?"

None whatsoever

For those readers who don't know what we are talking about, Babbage's envisioned Analytical Engine would have had about the same capabilities as a typical ordinary programmable calculator, such as the famous HP-42S (of which a free open source re-implementation is available as Free42). Of course the HP-42S was very very very much faster, but speed is not of the essence. But what is of the essence is that the HP-42 was also very very very much cheaper than the Analytical Engine, and it was mass produced.

There are literally thousands of programs available for the HP-42, and any of those could, in principle, be implemented on a functional Analytical Engine.

But the competitors of the Analytical Engine were not mysterious programmable calculators from the future, but human computers. The Analytical Engine was to be a mechanical programmable calculator, with ordinary mechanical speed; certainly not any faster than 100 human computers.

Let's try to understand the economics.

In the 1850, 50 real gold pounds sterling per year would have been a fantastic wage for a human computer; that would be about 7000 debased paper pounds sterling of 2022. Mr. Babbage burned through more than 17,000 pounds sterling (about 2,500,000 British pounds in 2022 money) before Her Majesty's Treasury pulled the plug. With that money the Treasury could have employed a hundred human computers for three years, computing whatever they wanted to be computed. With the advantage that the human computers did not require any huge capital investment, but rather they would have been pay-as-you-go.

The point is that one fantastically expensive, slow-as-molasses Analytical Engine would not have provided the British Empire with any competitive advantage. It could not do anything that a team of human computers could not do, it would not have been any faster, and it would have costed a lot more.

Meanwhile in America

It may be instructive to compare Babbage's visionary but utterly impractical dream with the severely limited but eminently practicable inventions of Herman Hollerith, who concentrated on designing and building electromechanical tabulating machines actually useful in real life.

Hollerith's machines were first used to tabulate the results of the 1890 U.S. Census; while the results of the 1880 census took eight years to tabulate, the results of the 1890 census were completed in six years, although the population had increased by 25%. Hollerith and other inventors continued to develop and refine unit record equipment, which became ubiquitous in commercial and governmental organizations throughout the first half of the 20th century; general-purpose digital computers only replaced them in the late 1960s or early 1970s.

In 1911, Hollerith's Tabulating Machines Company joined four other companies to form the Computing-Tabulating-Recording Company, which, in 1924, changed its name to International Business Machines, better known as IBM.

$\endgroup$
9
  • 4
    $\begingroup$ This brings to mind the current day issue of "quantum supremacy", where quantum computers today are not powerful enough - yet - to meaningfully compete with digital computers. That doesn't mean they won't lead to something better later on, but it will need much R&D to get there. Babbage's machine was similar: a very early example of programmable machines, but not quite useful yet. $\endgroup$
    – Brianorca
    May 24, 2022 at 16:09
  • 1
    $\begingroup$ Human computers sucked. You'd have them all work on a book of log tables, and every 20th number in it was wrong. Machines would have brought the sort of reliability to this that humans couldn't. They work 24/7 (minus downtime... and this would have had much lower downtime than tubes or relays). And the AE could have executed any Turing-complete algo, not just tables of numbers. It was a near-miss at having the computing revolution arrive 100 years earlier than it did. $\endgroup$
    – John O
    May 24, 2022 at 16:14
  • 2
    $\begingroup$ @JohnO: That you don't know how computing was done in the days that computers were young women does not mean that nobody does. No, it was not a near miss. It was just one failed attempt. As I said, elsewhere other people produced practical machines, even if (much) less ambitious. Electromechanical unit record equipment became widely available after the first world war, and revolutionized the way large businesses were run. (Electro)mechanical calculators were made and used in large numbers from the 1910s to the 1960s. $\endgroup$
    – AlexP
    May 24, 2022 at 16:41
  • 1
    $\begingroup$ @JohnO: "No one was teaching women to do computing in Victorian England": Exceptional women such as Mary Edwards made their living from computing long before the Victorian age. By the second half of the Victorian age women were fully accepted in the occupation. At the end of the reign of Queen Victoria, computing was well on the way to becoming female-dominated. (Women are definitely at least as good as men at it, and they accepted lower wages. Win-win, from the point of view of the capitalists.) $\endgroup$
    – AlexP
    May 24, 2022 at 19:29
  • 2
    $\begingroup$ Doesn't this ignore the fallacy of sunk cost and the difficulty of doing something the first time? Say Victoria herself decided cost was no object for the first one; who knows what fraction of the cost the second or third may have taken. They'd never be mass produced and might only be useful for academics, but still... $\endgroup$
    – user86462
    May 25, 2022 at 1:04
8
$\begingroup$

As stated on the Wikipedia page for the Analytical Engine

Babbage developed some two dozen programs for the Analytical Engine between 1837 and 1840, and one program later. These programs treat polynomials, iterative formulas, Gaussian elimination, and Bernoulli numbers

Gaussian elimination can be used to invert matrices, that is for example in solving systems of linear equations.

One possible application that comes to mind is the building design, where solving the frame requires solving such systems.

Babbage himself foresaw the development of Computer Science after the usage of his machine:

Babbage understood that the existence of an automatic computer would kindle interest in the field now known as algorithmic efficiency, writing in his Passages from the Life of a Philosopher, "As soon as an Analytical Engine exists, it will necessarily guide the future course of the science. Whenever any result is sought by its aid, the question will then arise—By what course of calculation can these results be arrived at by the machine in the shortest time?"[

$\endgroup$
4
$\begingroup$

Almanacs

While chronometers help with determining longitude for navigation, they must be recalibrated fairly frequently. The moon, planets, and their moons are excellent natural clocks, but their motions don't perfectly conform to Kepler's laws. Corrections for their gravitational interaction were well understood in the 19th century, but the calculations are difficult. An almanac giving the positions of celestial objects versus time was very useful to navigators. Even in the 21st century, almanacs are important official publications. Automating the calculations allows the use of more accurate models and has improved the reliability of the predictions. Navigation was, of course, a critical activity for a worldwide empire funded by oceanic trade.

Firing Tables

Firing tables are used to estimate artillery range given elevation, charge, projectile type, etc. Tedious to calculate, they are of obvious use to an empire.

Tide Prediction

Later in the 19th century, Lord Kelvin invented a special-purpose analog computer for predicting tides. That was, of course, very useful to the Empire. A general-purpose digital computer could have done this, and Babbage's work was a couple of decades earlier.

$\endgroup$
1
  • 1
    $\begingroup$ The artillery of the time wasn't precise enough for firing tables to be much use. $\endgroup$
    – Mark
    May 25, 2022 at 0:53
2
$\begingroup$

It creates a self perpetuating computing R&D race

Sure, the Analytical Engine is a bit of a toy. But now everyone knows that switches can be used to create thinking machines. Every artillery expert realises that faster computers = more accurate ballistics. Just having one computer incentivises the development of every technology that feeds into them.

As soon as relays get invented, they immediately get optimised for computers. People start trying to make vacuum tubes 40 years earlier. WW1 features WW2 style encryption and WW2 has semiconductors.

Development of the sciences, especially the social sciences

This doesn't scream competitive advantage, but what happens when Keynes comes along with the General Theory of Inflation and there's 80 years of computers being used to calculate things? Or when pervy old Kinsey shows up but the rates of sexual normbreaking is already known and tidily tabulated (and kills the scandal, as he's immediately disregarded as wrong)? Maybe the Soviets actually manage their economy quite a bit better with the aid of computers. The possibilities are endless.

$\endgroup$
0
$\begingroup$

Friction and slippage

No matter how well-oiled, there will be some resistance when turning the cogs. With a long enough chain of cogs, it may become impossible to turn the cogs without damaging them. This was a major problem for Babbage's actual Analytical Engine.

Cogs don't fit exactly together, or they wouldn't be able to move. If you have a series of cogs driving each other, the last cog will be behind the first by some small, but significant amount. This introduces imprecision, and the longer the chain of cogs, the greater imprecision.

$\endgroup$
2
  • 4
    $\begingroup$ I think the OP is asking "what problems can the machine solve?", not "what problems can the machine experience?" $\endgroup$
    – L.Dutch
    May 24, 2022 at 13:29
  • 1
    $\begingroup$ While L.Dutch is right that this does not directly answer the question, it is worth exploring more. We might assume that all of the same problems binary computing runs into could also be solved by this machine, but it can't. Ultimately, the machine has to compete with human computers as AlexP says, but it cant do that if it's answers are always a bit off. $\endgroup$
    – Nosajimiki
    May 24, 2022 at 14:33

Not the answer you're looking for? Browse other questions tagged .