Could the ancient Romans have built a digital computer?

I think the point of the article was "Could Romans build a computer, if plans were magically given to them?". Given this, the critique I am about to make is a little outside the scope of this thought experiment. However, I've wondered about questions like this a lot before, and I think it is also interesting to consider that the Romans (possibly) had many conceptual hurdles to cross as well. I am not a historian, but it seems that given their numeral system, they would not have had a firm understanding of mathematical bases, and thus would have not understood the importance of diodes/transistors/etc since they would have had to make the conceptual leap of representing numbers (and all information!) in binary (base 2). It's interesting to think about what obvious inventions we may not realize today due to the way we represent certain forms of information.

I think that mechanical computers are dismissed too quickly. Babbage was a perfectionist, and his machines were crippled by the reliance on decimal numbers. Konrad Zuse constructed his Z1[1] singlehandedly over a period of several years, based on a binary design. While the Z1 was only Turing Complete by technicality (it had the same instruction set as the relay-based Z3[2]) and it was fairly unreliable, it could have been built with more primitive technology and paved the way for significant improvements.<p><pre><code> [1] http://en.wikipedia.org/wiki/Z1_(computer) [2] http://www.zib.de/zuse/Inhalt/Kommentare/Html/0684/universal2.html</code></pre>

They absolutely could have built a digital computer and much more. He's slightly underestimating Roman enginuity for delicate manufacturing. Consider for example the Cage cup[1]. It's an exquisite piece of art and creativity and, all things cosidered, engineering as well due to the very tight tolerences and detail.<p>There's a lot more to progress than just the capacity for great leaps in engineering and sciences. Often times, there's a bit of luck and timing (Newton's apple and, of course, Newton himself) and also a societal willingness to promote new kinds of thinking.<p>The Antikythera mechanism is thought to be more a Greek invention than Roman[2]. And I've felt the Greeks were more appreciative of experimentation than the Romans who were driven more by the necesities of war and maintenance of the empire and they may be better candidates (with perhaps some Roman help) to go about making a proper computer.<p>We also need to understand that certain concepts in mathematics didn't exist in Roman mathematics at the time, like Zero (before Ptolemy)[3], which in addition to meaning "nothing" as quantity it also means "something" as a value.<p>We also know the Romans traded extensively in the Middle East which means they could have potentially been exposed to very early experiments with electricity[4]. While the age of "Bagdad Battery" like devices are debatable, it's not too much of a stretch to see ancient alchemists experimenting... more than likely the Greeks.<p><pre><code> [1] http://en.wikipedia.org/wiki/Cage_cup [2] http://en.wikipedia.org/wiki/Antikythera_wreck [3] http://en.wikipedia.org/wiki/0_(number) [4] http://en.wikipedia.org/wiki/Baghdad_Battery</code></pre>

A smarty pants answer would be, yes, but it took 2000 years....<p>The problem with this is that you end up with lots of additions to the premise. So, you ask the question: "Could Romans build a computer, if plans were magically given to them?. Then, you have to say, if they had plans they understood, then if they had plans using a certain technology, then if they had zero, and if they had an understanding of x, then y, then z, and on it goes. In then end, it becomes pretty clear the answer is no because of the sheer number of caveats we have to add to make the original premise work.<p>The whole thing seems to be a bit of an idolisation of an ancient civilisation. Its like we want them to have had mystical powers. Like how some people want to connect ancient Egyptians and Mayans to aliens. This is the fantasy of forgotten super knowledge.<p>Or, consider this: do we seriously that we could build something that we designed and built 2000 years from now? Something that could reasonably include or be based on alien technology and knowledge.<p>A connected tangent type thought, and I have no idea where it goes; but how come Romans were so clever, but current tribes around the world, sort of, aren't? How does that work? Is it a resource thing? Environment thing? Seems a bit odd to me that even now we have societies who are actually behind a society 2000 years old.

Why would they need to build a digital computer? How about an analog one. A good example is Moniac <a href="http://en.wikipedia.org/wiki/MONIAC_Computer" rel="nofollow">http://en.wikipedia.org/wiki/MONIAC_Computer</a><p>It uses water to model an economy. Taps can be used to change variables (eg interest rates). If the Romans wanted to do engineering calculations then there is no requirement for digital - an analog computer could do the calculations just fine. You can imagine taps to set lengths, widths, heights etc.

The problem really is in defining a computer. The first mechanical computer was not designed by Charles Babbage. Some thing like a clock/wrist watch is a mechanical computer. A marine chronometer like this(<a href="http://en.wikipedia.org/wiki/Marine_chronometer" rel="nofollow">http://en.wikipedia.org/wiki/Marine_chronometer</a>) is an advanced mechanical computer. The marine chronometer of the 17th century in our terms was basically a GPS device of that era.<p>Think of it that way. How do you think future generations hundreds of years now will look at our server farms. I bet our server farms won't even fall in their definition of a computer(or whatever term used to define a computer at their time).

<a href="http://news.ycombinator.com/item?id=2670203" rel="nofollow">http://news.ycombinator.com/item?id=2670203</a> has good links on mechanical computers and small CPU designs. <a href="http://lists.canonical.org/pipermail/kragen-tol/2010-June/000919.html" rel="nofollow">http://lists.canonical.org/pipermail/kragen-tol/2010-June/00...</a> concludes "There’s a strong appeal in the idea that universal computation has been within the grasp of human materials and manufacturing technology since the invention of sewing in the Paleolithic; it is only the mathematical sophistication that was absent. I don’t yet know if it’s true."

I don't understand the focus on tolerance for gears.<p>Tolerance is about miniturisation and commoditisation. These are not critical-path concerns here.<p>You could build a 'mill' analytical engine from very large wooden spars and wheels and it'd still be a mechanical digital computer.<p>The bigger question is <i>WHY</i> would they build one? What purpose would it solve?<p>Babbage wanted to reduce error in the lookup tables used by mariners. There wasn't a lot of other things it needed using for, and it would have been cheaper to instead have the charts compiled three independent times and comparing them.<p>What would the Romans have used a computer to compute?

The design of the difference engine and the analytical engine were dependent on tight tolerances in the parts, but that was a design <i>choice</i> not a fundamental constraint on mechanical computing devices. Indeed, look at "experiments" such as the tinkertoy computer (<a href="http://www.rci.rutgers.edu/~cfs/472_html/Intro/TinkertoyComputer/TinkerToy.html" rel="nofollow">http://www.rci.rutgers.edu/~cfs/472_html/Intro/TinkertoyComp...</a>) for examples of other designs based on other principles. The idea that the ancient Romans would have been incapable of building tinkertoy parts is simply not believable.<p>Also, people have managed to build difference engines using legos, meccano/erector sets. Here's a video: <a href="http://www.youtube.com/watch?v=KL_wy-CxBP8" rel="nofollow">http://www.youtube.com/watch?v=KL_wy-CxBP8</a>

This description does not discuss reliability. IIRC, these cat's whisker diodes need frequent readjustment. (<a href="http://en.wikipedia.org/wiki/Crystal_radio#Crystal_detector" rel="nofollow">http://en.wikipedia.org/wiki/Crystal_radio#Crystal_detector</a> agrees with that)<p>Using a single one to build a primitive radio receiver is easily doable, but combining a couple of thousands of them to get a 4004? I doubt one could ever get that to run for a second.<p>So, they would have to step up to better diodes. Of course, one can speculate that they could have done that, and they could. In the end, it is all down to the observation that flattery (aka imitation) is way easier than invention.

The somewhat recent 'Why didn't the Romans invent the internet?'<p><a href="http://news.ycombinator.com/item?id=5078275" rel="nofollow">http://news.ycombinator.com/item?id=5078275</a>

&#62; I don’t think the Romans could have built the analytical engine or other purely mechanical computer because of the tolerances required. I don’t know much about their manufacturing abilities, but I know they didn’t do a lot of it and imported most things.<p>But yet the greeks produced machines of incredible precision and function circa 150 to 100 BC.<p><a href="http://www.antikythera-mechanism.gr/" rel="nofollow">http://www.antikythera-mechanism.gr/</a><p>And whether they imported them or not, clearly the ability to manufacture to such precision did exist.

Not sure about that, but it always amazes me how the advanced the Greek and Roman societies were at the time, by the documents that reached our times.<p>If the fall of the empire were not to happen, who knows how advanced the mankind could be in our times.

Don't ignore the extremes. Before coming to conclusions with statements like "Romans didn't know X", "Romans couldn't do Y", reflect on what you mean: you are referring to the general and ignoring the rare extremes. I bet you could find a few people in the whole empire who knew about more advanced math and even the number zero. You could also find the best craftsmen to make intricate gears, and forge strong magnets. Indeed, that's how things would go in the old times. Masters would protect their secrets, and only if you had the money to hire them, you'd get the exquisite work. That's why, sometimes the know-how would be lost; like the examples for masterful crafts in the other posts.<p>In short, they didn't have the scientific and industrial revolution that we had. In their times, the masters were able to make money out of their knowledge and eliminate competition through secrecy. So, with enough money in hand, you could do a lot better than what was available to an ordinary person in the Roman times

&#62;<i>They even began dabbling in technology vastly ahead of their time. Hero of Alexandria drew up plans for a rudimentary steam engine in his Spiritalia seu Pneumatica.</i><p>Sorry, but Hero of Alexandria was a Greek. Not a Roman.<p>The Antikythera mechanism, mentioned casually in the article was Greek too.

I think the more interesting question is, <i>would</i> they?<p>Let's assume they can build the computer, including all the supporting technologies, like electricity and precise mechanics.<p>What use case would they have where it would make economic sense to spend countless man hours developing it?<p>Fortunately, all the component technologies are useful in themselves, so you could start by developing them independently, just like it happened naturally. Let's say you start with electricity. It's really useful right?<p>Well, without a power grid, electric engines, industry to use the engines, and large electric power plants to supply a steady supply of electricity, it is really much less attractive.<p>I'm doubting you could actually get a computer <i>built</i> until the civilization was ready for it, and there was a need.

interesting analysis, but I would prefer if it were pitched as - could a time traveler to ancient rome build a computer or could you build a computer while lost in the woods?<p>The article talks about the knowledge of the romans, regarding making wires and such, but building core memory or a transistor successfully and stably requires a very high level of knowledge of electricity, magnetism and some important higher level abstractions for thinking about storing information.<p>So, I think no, to a very high degree of certainty the romans could not have built a computer. If a society has not developed light bulb, I don't think they understand electricity well enough for a computer.

Romans didn't have a number zero.

In the Doctor Who episode The Fires of Pompeii, the romans in pompeii constructed a digital computer using blue prints given to be them by some aliens and doctor who destroyed it.

The broader question is why the technology of classical civilizations was not far better than it actually was, given that they clearly knew many of the basic principles - eg steam power. The traditional answer is "slavery": in societies where copious manpower, in the form of slaves, is so cheap (not free - you still have to feed 'em !), the incentive to productionise technology is vastly reduced.

I've always wondered about this. What are the chances electronic devices have been built in the past and for some reason there are no remains, we just can't tell them apart because we have no idea what to look for, or the materials they were made of have completely disintegrated by now?

The Baghdad Battery may have been invented back then, but there is no evidence that it is used for anything more than electroplating in small scales.<p>I think this article is actually quite interesting and given a bit more, could make for very compelling alternate history

&#62; I think if you traveled back in time to the Roman Empire and told them how to manufacture this stuff, you could plausibly create a very modest computer.<p>Then finally there'd be a use for the Pope's tweets in Latin.

So what would they do with this computer? (This pre-George Boole computer, so no Boolean algebra at the time).<p>Pre-Newton: rules out ballistics calculations, the favourite of early computers.<p>No Enigma machine or other calculating devices, pretty much rules out codebreaking.<p>No long term high capacity storage, or much in the way of large scale precise information available to work with.<p>Large, expensive, only one in existance, no lights for output. Nothing it could do that a dozen trained slaves couldn't do much quicker.<p>Maybe they could. Why would they?

Water power, someone else may have mentioned it, but water power was already used to power sawmills as possibly early as the 3rd century BC[1], I think a better question then isn’t why didn't they build a computer, as much as why didn't they use their water power to industrialise?<p>[1] <a href="http://en.wikipedia.org/wiki/List_of_Roman_watermills" rel="nofollow">http://en.wikipedia.org/wiki/List_of_Roman_watermills</a>

I think what was missing was theory - numeric systems, chemistry, electromagnetism and other sciences.<p>In the Roman era science wasn't invented yet - they were excellent craftsmen and sometimes Greeks produced excellent ideas, but there was also a lot of really bad ones and the facilities to weed out the bad ideas (peer review and scientific method) and propagate them (the printing press) weren't available yet.

Nice article, but I must take exception to one claim. The Antikythera mechanism is in fact a computer, it's just not a programmable one.

Romans building a digital computer? Seriously?? almost all of their so called achievements were actually stolen frm the East (India). 2500years ago India had two biggest universities in Nalanda and Thakshashila which encouraged scientific thinking. and you are talking about a barbaric civilization that was good at war and killing? pathetic...

For extra credit, figure out the surprising thing we can build now that won't otherwise be built until the year 3000.

Another way to read this article would be as follows: There has to be something <i>big</i> out there ready for <i>us</i> to build.<p>Now consider (just consider) that one does not necessarily need a "plan" or a "theory" to build things, one should definitely do so, preferably something (seemingly) useless.

The article is mainly a good reminder of how many things you can do yourself. The a posteriori reasoning without concrete proofs is faulty in the sense that the same reasoning can be applied to a lot of things beyond the romans: could the ancient greeks have discovered penicillin?

I like this because it's interesting to think about what <i>we</i> can build today with existing technologies that future civilizations will look back at in the same way we have done here! It's inspiring, all the potential around us that we cannot see.

I like to consider if I could build a computer if I was transported back to Ancient times. Having a fair amount of knowledge, it would still be an immense challenge to start from scratch and build up the tools and chemicals one needed.

How about a water computer?<p><a href="http://en.wikipedia.org/wiki/Water_integrator" rel="nofollow">http://en.wikipedia.org/wiki/Water_integrator</a><p><a href="http://en.wikipedia.org/wiki/MONIAC_Computer" rel="nofollow">http://en.wikipedia.org/wiki/MONIAC_Computer</a>

If they could, they would program in Lingua::Romana::Perligata.<p><a href="http://www.csse.monash.edu.au/~damian/papers/HTML/Perligata.html" rel="nofollow">http://www.csse.monash.edu.au/~damian/papers/HTML/Perligata....</a>

The Romans did something more masterful than building a computer. They built an empire with a system of roads, etc. without which, the Europe and the U.S. probably wouldn't exist today. But the incompetence of their emperors and the selfishness of their people eventually led to their demise. (They should have been friends with the huns, or have been able to adequates protect themselves.<p>Sure they could have built a computer if they had learned the technology. We are still amazed now as we learn of methods and technologies that human cultures had before us. Humans are much more able than we give them credit, and much more innovative than history endures and relates the resulting inventions.

This would be fun inserted into Rome, Sweet Rome:<p><a href="http://en.wikipedia.org/wiki/Rome_Sweet_Rome" rel="nofollow">http://en.wikipedia.org/wiki/Rome_Sweet_Rome</a>

&#62; While [the difference engine is] a beautiful piece of engineering, it’s actually not a computer.<p>What exactly is the distinction the author is making here?

Uh, in hindsight is so simple! Duh! Those lazy Romans.

According to my Civ 5 game they could.

I'm going to do my usual OT hijacking and mention the upcoming novel of a friend. Her premise was: how would an industrialised, polytheistic society cope with monotheistic terrorism?<p>It's an alternate history; the point of departure being that Archimedes of Syracuse was captured, taken to Rome and then funded (because the Romans were nothing if not practical). In her setting, the Romans are well into an industrial revolution.<p>The blurb she wrote for an agent:<p><pre><code> Pontius Pilate is a successful corporate lawyer headhunted by the Roman Senate to run a difficult province in need of a gentler, civilian touch. He’s been in the job five years — and is starting to get the hang of it — when the Yeshua Ben Yusuf file lands on his desk. This is not ideal right now. Judaea is in the midst of a major terrorism crisis, his wife keeps threatening to go back to Caesarea ( she can’t stand Jerusalem), his son is becoming far too friendly with the High Priest’s son, and his boss keeps forgetting that he isn’t actually in the army. Even worse, his closest friend and greatest rival from law school is Ben Yusuf’s lawyer. A Jerusalem courtroom is the setting for the first clash of civilisations, where people from a fundamentally different tradition are forced to engage with religious ideas that in many respects they do not want to understand. What is most distinctive about the book is my imagining of what a technologically advanced pagan civilisation would look like. That is, what if the Romans won much of their empire under conditions that we associate with the Industrial Revolution? What if — with their distinctive, non-Christian moral values — they were gifted with all that immense fire-power and confronted with monotheistic terrorists? I do not think the Romans were secular in the modern sense, and I haven’t portrayed them as such. They were, however, very different from the monotheistic peoples they confronted in Judaea. My Roman characters are still religious, but differently religious. Unlike many authors of a skeptical bent, I do not seek to score cheap shots by denigrating religion per se. Rather, Bring Laws &#38; Gods recognises the persistent vitality of religious traditions, especially when their practitioners are confronted with overwhelming military power and physical occupation by non- believers. </code></pre> It should be published this year and I am very much looking forward to reading it, based on the introduction and samples she's dropped:<p><a href="http://skepticlawyer.com.au/2010/09/22/the-past-is-a-foreign-country/" rel="nofollow">http://skepticlawyer.com.au/2010/09/22/the-past-is-a-foreign...</a><p><a href="http://skepticlawyer.com.au/2010/09/24/the-angel-bring-laws-and-gods-outtake-1/" rel="nofollow">http://skepticlawyer.com.au/2010/09/24/the-angel-bring-laws-...</a> (NSFW)<p><a href="http://skepticlawyer.com.au/2010/10/15/the-visit-bring-laws-and-gods-outtake-2/" rel="nofollow">http://skepticlawyer.com.au/2010/10/15/the-visit-bring-laws-...</a><p><a href="http://skepticlawyer.com.au/2010/11/25/patria-patria/" rel="nofollow">http://skepticlawyer.com.au/2010/11/25/patria-patria/</a> (NSFW)

What about electricity to use these things :) ? A schottky diode works on electricity...

tldr: maybe, with a big if around powering it

You do not need powerful magnets to make a generator. You use whatever permanent magnet you have and put an electromagnet in parallel with it. When it starts turning, the output power from the permanent magnet can be fed to the electromagnet and the generator pulls itself up by its own bootstraps.

I think the bigger question is: could they want to?

Yes, but it wouldn't have been able to run Ruby ; )<p><i>(smile, it's funny)</i>

This seems like a classic example of Betteridge's law of headlines...