Hacker News new | past | comments | ask | show | jobs | submit login
Why didn't the Romans invent the Internet? (shkspr.mobi)
57 points by edent on Jan 18, 2013 | hide | past | favorite | 62 comments

I think the author is showing a bias common in technology circles: assuming that the limiting factor is technological knowledge.

But history is full of examples of people who had all the necessary knowledge and yet didn't execute. The ancient Greeks actually built toy steam engines. So why didn't they launch an industrial revolution?

The limiting factors have much more to do with availability of capital than raw scientific and engineering knowledge. You need the vast web of people with specialized knowledge and tools that keep any technology running. Technology as simple as a pencil has a mind-boggling array of underlying capital requirements (see Leonard Read's classic essay "I, Pencil": http://www.econlib.org/library/Essays/rdPncl1.html)

Remember that even a skill like "reading & writing" made you an unusually valuable specialist in a pre-Gutenberg world.

The true capital costs of a Roman semaphore internet are almost impossible to guess. Given the myriad other demands on their capital, it's not at all surprising they didn't try it.

"But history is full of examples of people who had all the necessary knowledge and yet didn't execute."

Great point. Future observers might ask why our society doesn't exclusively use self-driving cars, renewable energy, all-electric vehicles, etc.

I take your point, but all three of your examples currently have severe physical limitations. It is not just a matter of execution. Untested, inadequate, and low range.

I agree that there are non-technological factors for why they didn't invent semaphore, but I seriously doubt that a lack of capital had anything to do with it. Have you seen the sort of things the Romans _did_ try? Stupendously grandiose and expensive things. Rome was awash in capital. A half-decent semaphore network would've been a rounding error in their budget.

Rome was awash in capital, but it was missing other economic structures that help support innovation, such as the corporation.

Even at the height of the Roman Empire, or later Italian Renaissance, the accumulations of capital that existed were somewhat hamstrung by the fact that they were concentrated in the hands of individuals and families. So personal aggrandizement was a huge motivating factor in deciding how the money would get spent, hence the large mount of art and monumental construction projects we see at those times.

I'd suggest that corporations are much better for producing technological innovation. As compared to what things were like in Rome, their key characteristic was that decisions about how money would get spent are generally made by groups of people who don't actually own the money themselves. That cuts way back on the motivation to just blow it all on yet another temple in one's own name.

This is a really excellent point, and I think cuts much closer to the truth than a simple lack of technology or capital.

If you think about the great infrastructure accomplishments of Rome: roads, aqueducts and sewer systems, rapid large-scale concrete construction, hypocausts, etc., then it seems to me that almost all of them were essentially scaled-up versions of domicile-level or urban-level technologies. These technologies could be piloted at a small scale and then scaled up if they proved to be worthwhile. The institutional mindset was almost totally civic rather than imperial. The military was something of an exception to this, and of course individual merchants and merchant companies would have been very much focused on keeping the gears of the empire turning. The roads and the granary systems were scaled-up urban technologies that helped the empire to function, but I'm not aware of any innovations at the empire level which didn't have some kind of urban precedent. Presumably there were no political or commercial institutions which were structured to support that kind of investment -- or even to ask those kind of questions.

The British thinking around empires was fundamentally very different. They genuinely thought AS an empire -- a highly distributed network rather than an overgrown city-state. This gave them the insight to support the creation of the semaphore system.

Perhaps a more interesting question, in my opinion, is why didn't the Chinese didn't invent the Internet? I'm not as knowledgeable about Chinese history, but my impression is that they were vastly better at thinking and acting like an empire -- not in the expansionist sense, but in the administrative sense. Presumably they would have had the kind of institutional basis to develop a network of semaphore towers. So what was the obstacle for them?

Fun fact: The Chinese invented a movable type printing press almost 500 years before Gutenberg, but it wasn't practical due to the enormous number of characters in Chinese.

Movable type has a worse story. Although not usually mentioned it existen in the late minoan age. Delayed for only about 3500 thousand years... [1]

[1] http://en.m.wikipedia.org/wiki/Phaistos_Disc

That's not why the Greeks didn't create an industrial revolution...

"The uselessness of the aeolipile is an excellent example of the vastly under-appreciated role of materials in technological development."

The British had centralized government planning of materials science research for use in cannons and other armaments. That paved the way too, not just a pure invisible hand of the market.

A lot of time passed between between the Greeks and British.

Near the end of the Roman Empire, the Romans faced a labor shortage and an important economic system was implemented by Constantine, which would later be copied by everyone else and dominate Western Europe until the Renaissance (It began to decline after the Bubonic plague improved the labor bargaining power of the less well off) . The system was called later called Feudalism, and abolished private ownership of land.

I'd add another point, the Romans are a better comparison to the Greeks, as they were closer in time. the Romans were far more centralized than the Greeks - The Athenians were directly democratic - That it was hard to scale so the Romans had an extensive system of laws.

In terms of overall scientific output there is no comparison, the Greeks were greater by a massive margin.

William Gibson refers to this cognitive leap with existing technologies as Steam Engine Time: http://www.williamgibsonbooks.com/archive/2003_01_28_archive...

yea i think the mindframe is a much huger factor than people realize. people underestimate just how different the world looks to people in different places/times/cultures

sidenote: pg wrote an essay (i forget which one) where he points out that blogs could have been around years earlier than they were. anyone remember which essay it was?

Well, they did.

If you travel around Europe you will see all the towers that Romans used to carry messages. Lots of them, specially in places like Spain or Greece with lots of mountains. In other places like most of France such a tower is far less useful as there are less differences in height.

Those towers were already being used before Romans took them, for example in the North of Spain they would be used to alert from Viking Invasions from sea, it will alert the "Castros". We are talking centuries before the Romans.

About complex messages being transmitted far away, look at the Silbo Gomero at the Canary Islands, it was used before the Spanish conquered the Islands in XVI century(we don't really now how much older than that was, but probably centuries): http://www.youtube.com/watch?v=qlZh9I1pxj0

for example in the North of Spain they would be used to alert from Viking Invasions from sea, it will alert the "Castros". We are talking centuries before the Romans.

... except the Vikings were well after the end of the Roman empire.

Well after the end of the Western Roman Empire - the Eastern Empire even had Vikings and Saxons fighting for it in the form of the Varangian Guard:


Well fine, but Byzantium never took Northern Spain, let alone 100's of years after the Viking age.

That's very interesting (especially the whistling language). I suppose what divides Roman watch towers from the Internet (from my POV) is that they could only carry a fixed messages. They weren't able to send general communications.

That said - the geography aspect is not to be overlooked.

The Romans (and many other civilisations) used to build beacons to signal various things along walls and coastal lines.

The Roman "internet" was "roads." The extensive network of roads allowed for much faster communication along with the ability to actually act on it with speed.

Maybe Rome was the first Amazon.com ;)

"If the Romans had built a practical semaphore, not only would we all be speaking Latin"

There is a reasonable argument that Spanish, French, Italian, et cetera are but dialects of Latin. It is only Euro-centrism which treats them as separate languages while dumping diverse tongues into Chinese.

> It is only Euro-centrism which treats them as separate languages while dumping diverse tongues into Chinese.

This couldn't be further from the truth; it is the Chinese government that insists that all different Chinese languages are merely `dialects' of a common Chinese tongue, based on a Soviet-era ideology of national unity.

China does recognize distinct languages like French, Italian and Spanish, because it recognizes France, Italy and Spain as independent nations. This view isn't inspired by Euro-centrism, but by the idea that China is an indivisible nation while Europe clearly is not.

> This view isn't inspired by Euro-centrism, but by the idea that China is an indivisible nation while Europe clearly is not.

Euro-centrism isn't the culprit here; it's the fixation on the arbitrary concept of the nation-state that's responsible.

The concept of the nation-state was created at the treaty of Westphalia to solve some very specific problems and is a lot better than what came before.

It's easy to knock things if you ignore why they were created or what existed before them. Also, you can't abolish something if you don't have anything to replace it with.

Who's knocking anything or looking to abolish anything?

Being aware of the limitations and biases of one's own paradigm and cognizant of the fact that it's an artificial construction is different from declaring that it has no value.

Not quite. I'm European. Most Chinese, like my wife, pride themselves on hailing from a nation of 1b+ people who share a commonly intelligible language - and rightly so in my opinion.

Could you talk some more about this? I've heard the counterargument about an overwhelming number of incompatible regional dialects too often.

The way I've heard it explained (and I only studied Mandarin for a year) is that a common language and writing can be simultaneously compatible and incompatible.

Consider these stereotypes - an urban youth from inner city America and a gruff old Yorkshireman. Both speak English and use the same alphabet and mostly the same words. Yet the syntax and slang differences each have could render the dialects meaningless to each other.

But, write down what they are saying and suddenly the meaning becomes a lot clearer. Yes, there are differences, but you can probably work your way around them.

Every language has is fair share of incomprehensible dialects. Yes Brits and Germans I am looking at you.

"A language is a dialect with an army."

A more accurate title is "why didn't the Romans invent the optical telegraph". The reason why the optical telegraph didn't become the Internet is very simple: bandwidth. You need lots of spare bandwidth before you can consider adding extra information to messages so as to create the various protocols that the Internet is built on (Ethernet, TCP, IP, HTML etc.). That's not something you're willing to do with a system that, even when perfected by the French in the 19th century, could only transmit about 1 symbol / minute [1].

[1] http://en.wikipedia.org/wiki/Optical_telegraph#France

"Why didn't the romans invent the internet?"

Simply put you could't beat the bandwidth a messenger, traveling on a transcontinental network of roads (http://en.wikipedia.org/wiki/Tabula_Peutingeriana), changing horses every few hours. Semaphore communication would've been a lot slower. Oh, and it wasn't any good for bringing things.

"Imagine what the world would be like if we'd had a 2,000 year head start on the principles of the Internet?"

Most likely the same it is today. See http://en.wikipedia.org/wiki/Dark_Ages for reference.

I think the article is built on the false premise that the existence of knowledge per se, guarantees it is going to be used. To paraphrase Kenneth Clark (http://en.wikipedia.org/wiki/Civilisation_(TV_series)) you need a flourishing civilization to utilize the knowledge you have. Western Europe effectively UNLEARNED Roman knowledge and didn't "discover" it again for a whole millenium. Although never forgotten, Eastern Europe (represented by the Byzantine Empire) didn't have the scale or resources to take advantage of this knowledge.

So there you have it, knowledge alone is not enough.

Bandwidth != latency.

That tabula Peutingeriana looks interesting -- thanks for the link.

There's a great early science fiction novel about the possible effect of semaphore communication technology on the Roman empire — L. Sprague de Camp's "Lest Darkness Fall", http://en.wikipedia.org/wiki/Lest_Darkness_Fall. In the novel, de Camp argues that a time traveler with this technology (and foreknowledge of history) could have prevented the Dark Ages.

It so happens that a blogger I host has written an alternate history novel set in a late-Republic / early-Empire Rome that has undergone an industrial revolution. She takes as her point of departure that Archimedes of Syracuse was captured, and not killed, by the Romans.


I really must pester her to finish editing and get it published.

Edit: she says it's being considered by a publisher at the moment, but may go under a psuedonym, so I've removed her name from this post.

> I really must pester her to finish editing and get it published.

Yes, you do, because I want to read it, and I know several others who would too.

She posted some of the stuff that was cut out of her draft:



Some of the first link is quite NSFW. Take that as you will.

The Romans were too busy inventing sanitation, medicine, education, viniculture, public order, irrigation, roads, fresh-water systems and public health.

So like others have said, infrastructure.

But apart from that, what have they ever done for us, eh?

> The Romans were too busy inventing […] medicine, education, viniculture

These three long predate romans:

* Imhotep is the first known physician (technically a polymath, but one of his specialties was medecine) in ~2600 BCE and we have his descriptions of a number of disease diagnosis and treatments

* There are traces of formal education dating back to >1000BCE in ancient India and China, scribal education networks in ancient Egypt and middle-east.

* As for viticulture, while it spread to Europe through Rome it predates Rome by millenia: http://en.wikipedia.org/wiki/History_of_wine http://en.wikipedia.org/wiki/Viticulture

Although a lot of things were "not quite novel" in those times (as are not nowadays), there were a lot of improvement and development to be done around them (something valid today, too).

Visual telegraph systems became very popular around the coast of the UK (particularly England) at a time when it was also possible to make money in the commodity markets by knowing which ships were approaching port (and by implication their cargo) before anyone else.

So one might suspect that rapid messaging systems became popular when there was an economic case for them.

Not only an economic case, but also being able to use your tech advantage without being bullied by an established competitor. "Oh you're using this weird contraption to outsmart me ? Cute, I'll use daggers in the hands of assassins, whatever bailiffs use to enforce trumped up court orders, or something of the kind." In other words, you need the protection of the law as an upstart.

It can be argued that this was missing in the Roman world, while the economic case existed and even the tech was known.

And you think at the start of the Industrial Revolution there was a perfect rule of law in UK & France? There were lots of thugs back then too.

> In other words, you need the protection of the law as an upstart.

Rome in the mid to late Republic and early Empire had very sophisticated and stable laws and court systems, including a body of commercial law that was not seen again until thousands of years later.

A very reasonable point. Although the Romans had a highly mercantile system relying on foreign trade.

The lack of a proto-internet by the Greeks and or Romans was as much a sociocultural issue as it was technological.

The author seems to ignore the basic requirements for the technology to exist to begin with (a mistake Orwell made in 1984 for example). It is my opinion that an unfree society would not have invented the Internet.

Although it's hard to argue about what a theoretical unfree society would have accomplished, we know the USSR had the M-40 system in the early 1950s. That was a full duplex radio data network for missile defence. Whether that would have morphed into the civilian Internet in the same way ARPANET did is impossible to say.

Which basic technology - other than the telescope - was missing from the Roman's knowledge that would have enabled a proto-Inernet?

Where does Orwell suggest anything like the Internet in 1984?

His 1984 does have quite a lot of high technology (machine generated media content, telescreens) but nothing that I would see as being incompatible with a totalitarian state.

"The Great Wall of China had a network of high towers where news of attacks or raids could be passed down the wall by means of lighting a signal fire" http://www.wisegeek.com/what-is-a-signal-tower.htm

That was also in the Europe too, both before Romans and of course also in their times. Current time Romania's Transylvania for example (Dacia in the Roman's time), is one of the places that was situated at the edge of the Roman Empire. Because of the risk of eastern invasions (migration from east happened all the time), it had an efficient system of fire/smoke outposts signaling the presence of possible enemies over Carpathians, long outside the Roman territory.

Interesting. I hadn't considered the Great Wall. I wonder how often (if at all) it branched? Sending a message down a line should - eventually - lead to sending messages through a network.

That`s exactly what I thought of when reading the article and I expected it to be mentioned at the end. But I only could find your comment here. Well done.

I recommend reading

Data Communications - The First 2500 Years G.J. Holzmann Proc. IFIP World Computer Congress, Hamburg Germany, August 1994, (Invited paper). http://spinroot.com/gerard/pdf/hamburg94b.pdf

The book:

The Early History of Data Networks G.J. Holzmann and B. Pehrson IEEE Computer Society Press, Los Alamitos CA, (John Wiley & Sons), October 1994.

The basic ideas of relaying, encoding, separation of control/data information, error control, flow and rate control go back surprisingly far.

An interesting read on the subject, very detailed and it goes as far as the first letters and the beginning of writing: The Information by James Gleick. [1] In a chapter, he explains how the transmission of messages worked with towers, the limitation they had, the different things they tried. For instance, he explains how they used a cross with moving branch and each position of the branches would mean a letter, but also how it was difficult to implement and how it eventually evolved.

[1] http://en.wikipedia.org/wiki/The_Information:_A_History,_a_T...

I'm guessing the reliability of the message transfer, and the consistency of its contents declined with the increase of distance. It's like the kids game where you whisper something into someone's ear, which is passed along the same way along the line, and you find the end-receiver wound up with a completely different message.

That's probably why they sent our messengers with wax-sealed letters. If they were to build and use such an important infrastructure, it would surely mainly be used for military/political purposes. The criticality of those messages would inherently demand that the message would be delivered in an unaltered state.

"it would surely mainly be used for military/political purposes"

in the same way that roads were built for military purposes, and the Arpanet was built for military purposes?

In a thousand years someone will speculate about why the people of today didn't just build and exploit [x]. Stop messing around and go build [x].

It seems a more plausible invention in the Hellenistic kingdoms Rome disrupted and conquered. They made expensive and farsighted infrastructure innovations like the library and lighthouse of Alexandria. (The lighthouse beacon could be seen from ships 30 miles out, according to The Forgotten Revolution.)

For anyone more interested in this topic, read "The Information", fascinating book.


If only...

Give examples.

If only...wifi hardware manufacturers included ad-hoc network configuration an aggressive default. To wit: any time another device was in range, spare bandwidth would be applied to augmenting regular planned network connections with an on-the-fly spontaneous network. Various complaints could be addressed with ease on a per case basis, but point is the norm would be applying unused capacity.

Wasn't included initially because there wasn't enough expected ubiquity to be feasible. Now that everything but my toaster has wifi support, we could have something approaching a robust free universal Internet not beholden to backbones and pricy data plans.

Wifi's range is too short to make a large mesh network genuinely feasible. Even in densely populated cities you're going to have gaps where you need to cross a street. That'd require directional antennae and a huge amount of deliberate effort and maintenance to hook up a large area.

Must have longer omnidirectional range and some degree of wall penetration. I'm not sure if that's possible within reasonable FCC limits of frequency and power.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact