Did we build the ‘right’ Internet?

An interview with Prof Andrew Russell

The longer I have been in the tech industry, the more I have come to appreciate the hidden complexity and subtlety of its past. A book that caught my attention is Open Standards and the Digital Age by Prof Andrew Russell of Stevens Institute of Technology in New Jersey. This important work shines a fresh light on the process that resulted in today’s Internet.

For me, it places the standard ‘triumphant’ narrative of the rise of TCP/IP into a more nuanced context. It is too strong to say that this work debunks that oft-told tale, but it certainly reframes it. I have interviewed Prof Russell, who has kindly shared his many insights.

MG: What inspired you to write this book?

AR: I am a professional historian, and my book brings together two streams: the history of the Internet, and the history of engineering standards. Historians need to know more about the Internet; conversely, Internet people need to know more about its history.

Before I went to graduate school in 1999, I was working at Harvard on a project on information infrastructure. We were asking ourselves “what is ‘cyberspace’?”. We knew this phenomenon was a big deal, and there was a great deal of interest from faculty and researchers in many departments at Harvard and MIT, but I didn’t see any historians involved.

So I decided to study the history of the Internet. What I found was a narrow focus as portrayed exclusively by insiders. For instance, organisations such as the Internet Society had captured a blow by blow technical history that flattered them.

I pursued my MA in History at the University of Colorado, where I took courses in intellectual and cultural history, as well as courses about standards and public policy at Colorado Law School, in the company of Tim Schoechle, Ken Krechmer, Phil Weiser, and Dale Hatfield. As a result, I began to see the field in broader terms.

The real story of the Internet is not technical, but human. It is about collaboration amongst engineers, and how that is always in tension with competition (especially in an American setting). It also has a dimension of international collaboration, as previously seen in telegraphy and telephony. The human legacy of the Bell System also plays a crucial role.

For my dissertation I studied standards in communication networks in various settings, including the transition from 1G to 2G wireless, the emergence of the Internet, and the roles of both government and private sector in these developments. From this work, I began to form an important conclusion about the American techno-political economy in the 21st century.

A key industrial change wrought by computing is the emergence of consensus-based engineering standards. This contrasts with telegraphy and telephony, where most standards had been decided by monopoly diktat. The origin of consensus standards is not within communication networks, but instead within fields such as electrical and mechanical engineering. Computer networks had forcefully brought these worlds together, generating a new lexicon and mind-set of openness in the process.

Why the focus on open standards, given their potentially dry and technical nature? What makes these so figural to a historian?

People in my field aren’t scared of dry and technical topics; the challenge is to show why they matter. It’s only dry and technical when it’s in the hands of technologists and engineers. As historians we study people. Standards are important to the people we study, therefore they become important to us.

With the emergence of the Internet this term ‘openness’ became an increasingly prominent buzzword. The standards people seized on it in the late 1970s, which is earlier than what we might expect, and certainly earlier than existing histories suggest. The term straddled diverse domains: strategy, culture, economics, etc. So what did it mean?

The concept of ‘openness’ has a long and rich context. For instance, the themes of an ‘open society’ with an ‘open door’ reflected American foreign policy at the end of the 19th century into the early 20th. These in turn reflected the nature of American engagement with Asia.

The definition of that role was no longer primarily militaristic, but instead was based on trade, an activity that American diplomats framed as mutually beneficial. Of course, the Americans knew that they would benefit more from an economic engagement than from a military engagement, and wanted to avoid confrontation with the European empires who had footholds in Asia.

This idea of ‘openness’ also has a very broad cultural appeal, feeling both welcoming and warm. In contrast, ‘closed’ does not feel positive. To illustrate this, consider the book “The Closed World” by Paul Edwards. He documents the story of Cold War computers in America. The intelligence community had a fantasy that they could model everything, in Dr Strangelove style.

So our framing is a transition. It starts from a closed world with a seed of openness, and this persists throughout the Cold War period. The idea of ‘openness’ in the context of networking then emerges in the 1970s and 80s. By the 1990s ‘free’ and ‘open’ are used regularly in civil liberties terms.

This transition really jumped out to me as a significant change in use of language, and a jump from earlier contexts. We see accumulated meanings, sometimes contradictory, but generally appealing to the ideals of those living in a liberal democracy. The work of Chris Kelty in “Two Bits” had captured the history of ‘free’ software and the meaning of ‘freedom’. As he had taken ‘free’, I grabbed ‘open’!

The concept of being ‘open’ is fundamentally a relational one. The question I began to research was “who started using it?”. The earliest mention of ‘open standards’ that I could locate was the British computer expert Jack Houldsworth, a leader in the British proposal in 1977 to ISO to create a new standards committee for computer network standards. The reasons why computer people started to use this word ‘openness’ are hazy and not well documented.

We can conclude that ‘open’ had acquired significant meanings by the 1970s that gave both strategic and PR advantages in spreading ideas. Hence the term caught on. Similarly, ‘internet’ caught on, whereas ‘catenet’ did not, as the former had a broader base of appeal. The term ‘modular’ likewise has a fascinating intellectual history. Sometimes you can nail it down, sometimes it’s evasive. These quests are part of a wider genre called ‘conceptual history’.

Henry Ford famously exclaimed that “history is more or less bunk.” How is networking history (more or less) bunk?

Layers and layers of myth and misunderstanding are accumulating by the day.

First are the intentional or ideological misinterpretations. That Al Gore ‘invented’ the internet is a deliberate (and clumsy) attempt to obfuscate the historical record. Likewise, some claims to the ‘invention’ of email are clearly bunk. Such confusions in the popular imagination often arise from mixing up the invention, naming, and patenting of a thing. These are issues that my friends in media studies try to clarify, but ignorance is a powerful force.

When Steven Johnson wrote about Internet history in the New York Times – saying “we built that” – he twists the record. It takes credit away from the US Department of Defense for seeing the potential of data networking. Similarly, when ideologues in the Wall Street Journal insist “Xerox built it” (or any other corporation) it mangles the historical record by wrongly claiming the Internet was VC funded.

Then there is the usual “winners write history” effect, which is very strong here. This is subtler and rather interesting. For instance, there are dozens of interviews with Vint Cerf about the history of the Internet. Cerf has done a good job of honestly representing his role. But the Internet doesn’t have a single inventor, as it’s not like the light bulb.

Hundreds of people had roles, many with aspects of computer networking that didn’t work out. So they went on to do other things. Nobody with a 30-40 year career wants to focus on those things that they did decades ago that didn’t work out. This success filter biases the historical record.

We also like our stories to have heroes, and ARPA sometimes takes up the hero’s role. As Mariana Mazzucato has documented in her book “The Entrepreneurial State: debunking public vs. private sector myths”, the government has had a foundational role in many of the component technologies, with that effort spread over many programmes and agencies.

Also, there is a normal progression of histories of recent things, whereby several generations unfold, each of which is increasingly critical and contextual. If we take other technology histories as our guide, we see that it would be bizarre if the narratives told by a technology’s inventors persisted as authoritative for more than a decade or two. As technology matures, some things change in unexpected ways. The histories also should change and will change.

For instance, the Internet isn’t proving to be as robust as it seemed 15-20 years ago. We see security issues, and variable performance, and phishing scams, and terrorism. As these problems are coming to light, it brings people to take another look at this thing we have been adoring. We are made to ask ourselves if we should adore it quite so much.

Taking security as an example, Edward Snowden prompted a lot of people to think about fixing it. There have been calls for the IETF to step up, but they just don’t have these capabilities and cannot rewrite the past. And they cannot overcome the technological and bureaucratic momentum that they created for themselves.

So this evolving story may not be a happy one of liberation and freedom, like the 1990s claim, but rather may be a sad tale of a new utopia and its dreams slowly being dashed. We are being left with a dystopian and authoritarian cybermall, and this is happening in front of our eyes. It is a generational effect, whereby we get used to the technology and experience its unintended consequences.

Then there is the re-telling of stories. With the the American Revolution, for example, we have had constant reinterpretation. That narrative arc starts with the march of freedom, followed by the Founders backlash, making them the slave-owning bad guys. In the 20th century we saw economic interpretations, where the focus was not freedom, but autonomy over property. We’re still working on understanding the fall of the Roman Empire!

Finally, the story is inherently complex. It lacks clear and charismatic heroes and heroines. It doesn’t have any obvious moral clarity or ambiguity. Hence networking history is not an easy story to reduce to clever or popular slogans.

Where is the ‘standard’ history of the Internet weakest, in your opinion?

If there is one specific way networking history is bunk, it is that it lacks awareness of wider and richer possibilities. Engineers making that history inherently live in a time of limited possibilities. We need to expand our awareness of ‘might have been’ and ‘could yet be’, since that is the only way to make something different and better.

As a case in point, where is the science that underpins these networking endeavours? It would never had occurred to me that the science is missing if I hadn’t met John Day, author of “Patterns in Network Architecture” and inventor of Recursive Internet Architecture (RINA).

This touches on a more general and important issue of the relationship between science and technology. The misconception is that “technology is merely applied science”. This has not been true for most of human history. Natural philosophers and craftsmen have always been different. My Stevens colleague James McClellan and the late Harold Dorn wrote the definitive account of this story in their book “Science and Technology in World History” – just out in its 3rd edition from Johns Hopkins University Press.

Yes, we see from the late 19th century onwards how scientific advances have triggered technology changes. But they do not necessarily go hand in hand. Nor do they have a linear relationship, as proposed by the standard rationale for investment in basic science.

Indeed, as we have seen with the Internet, we are capable of building sophisticated machines that do great things, yet without fully understanding their mathematical or scientific basis. What we have done, or could yet do, is not fully visible to us.

This is not an isolated example. We built aircraft before we mastered aerodynamics, and steam engines before the science of gases. The story of chemistry’s emergence from alchemy shows how it took a long time to understand what elements are. The periodic table is an incredible breakthrough – but one that is after the fact.

The domain I explore is the relationship of science and technology with world history. What you have to appreciate is that science and technology are distinct spheres that sometimes interact, rather like literature and fashion.

These human role issues touch on questions of social class and the related status and privilege of science over technology. This can also be seen in corporate structures, whereby we celebrate the intellectual endeavours of Bell Labs over the more pragmatic achievements of the machine shop at Western Electric.

What do you know now (that you didn’t know when you started this project) and wish more people also knew?

Looking at the history of the Internet, what strikes me is just how so unlikely it all is, and that it has unfolded this way. We spend so much time creating a ‘clean’ narrative to tell to students, or our relatives over a holiday meal. Doing that cleanses the story of all the dead ends, coincidences, and contingencies of the past.

So I have been feeling around to expand the story beyond the ‘too clean’ narrative of the ARPANET turning into the Internet. There were so many more people working on computer networks and related technologies. Many didn’t know what there were doing, or stumbled into this field from adjacent problems. My appreciation for that has grown.

In doing so, I faced the challenge of how would I write a new book offering an alternative history of the Internet. I felt it would be excruciating to leave out many things, and became intensely aware of what I don’t know in this process, which was basically ‘everything’. The certainty we have that we know the story is based on a collective false consciousness.

Again, take the story of John Day. I heard tales of his ‘sour grapes’ that his ideas didn’t get adopted. Then I looked into it and quickly concluded “Good God! Whatever happened!”. I wish more people knew it could have been different, and in a good way. It wasn’t just a competition between ideas based on their technical merit.

This is a part that is truly bizarre to me – although with many precedents in the history of science and technology. The conventional tale is that the Internet won out because it was the best technology. In fact, it wasn’t the best technology; it won out for social and political reasons, mostly due to the financial power of its sponsor in the American Department of Defense.

It is also a bit terrifying that this might be true for almost every sophisticated technical system. Flaws that can only been understood by someone with a deep technical knowledge get written out of the story.

Who do you feel are the unsung heroes (or unseen villains) of open standards and the Internet and why?

There is unquestionably one unsung hero, who I was able to interview: Marc Levilion of IBM France, who is very bright and skilled technically. He devoted his entire career to making these things work. He had a big role in IBM in France, as well as the ISO. Being smart and doing the right thing does not mean you will be rewarded for it.

He had to argue against himself to represent IBM’s position faithfully. He was trapped in a bureaucratic organisation whose forces squeezed out the rational solution he was personally advocating. This harmed his career and many people like him. The history of technology, computing and the Internet is full of people getting punished or not getting recognition. The transcript of my interview with him is available from the website of the Charles Babbage Institute, in their oral history collections.

There are many, many, many more like him. Take Lisa Rajchel and Tilly Bayard-Richard from the ANSI and AFNOR/ISO secretariat. Their hard work has also not been recognised. We like to talk about the dreamers – the Englebarts of the domain – but they are only a small slice of the story. If we want a rich understanding of this history, we need to appreciate and celebrate different types of labour that bring new technologies to fruition.

Another example is the exceptional and charismatic Louis Pouzin, along with his delightful Cyclades team, still loyal from their formative experience together in the 1970s. His public reputation recently has been rehabilitated. He was doing smart and rational things in France with Cyclades and got punished for it. It’s fun to meet them, since they were doing spectacular things with a gusto and spirit anyone would admire.

It is just that, for whatever reason, it didn’t turn out to be the winning hand from a historical point of view. The team was broken up, and the project was cancelled. Louis had to be coaxed into talking. I was only able to make contact with him and his team thanks to the diligent efforts of Valerie Schafer, who is the leading French historian of computer networks. When I spoke to one member of Pouzin’s team, Jean-Louis Grange, he exclaimed that “This is ancient history! Why are you asking these questions?”.

John Day was in a similar position, but instead he has stuck with it. Where so many people involved with non-Internet efforts in the 1970s and 1980s simply moved on, the richness of the problems have held John’s mind captive as he went off to work at different companies. His ‘big idea’ about packet switched networks got hold of him and he never let go. He has devoted so much energy and time to it, and is a real scientist in a field of builders and makers and sellers. He keeps on pushing and pushing to understand some foundational aspects of the problem space. In that respect he is exceptional.

When it comes to villains, it tends to be a question of “whose reputation is overblown?”. We have seen a long conflict between people who got the lion’s share of the credit: Kleinrock, Cerf, Kahn, etc. But it was a group effort involving hundreds of other people, who have moved on, or passed on. So a detente is now emerging – or so it seems to me, although my sense is that there will always be tension among some individuals.

I have made a point to try not to get too close to any of them personally, for better or for worse, mostly because I have no desire to referee between competing egos or competing claims. I don’t have much interest in that aspect of this history. I prefer to focus on matters of structure and process, on institutions and ideas, and on technology and culture. So I am happy to leave personal disputes to biographers and journalists.

If there are true villains, they are the national security state and the self-interested capitalists. They only want to increase their status and power, and often misrepresent the importance of their role and value in the development of the Internet. But I think if we looked closely even at these ‘villains’, we would see simple humans, with human flaws, each pursuing a certain set of interests for reasons that made sense to them, no better or worse than the rest of us.

Looking at the grand sweep of history, how might future generations perceive the emergence of the early Internet?

I wish I knew! Triumph or tragedy? Is the free and open Internet an aberration, or the new norm? As a historian, the future is outside my zone of expertise. Is it a great flourishing of the human spirit taking shape in a ‘global brain’ or ‘global village’? A new era of human understanding? Or dystopian ‘Skynet’? Or something in the middle?

I just hope the Internet doesn’t turn out to be too big of a disappointment. It has unleashed many new and cool things. It is so optimistic by its nature that I hope it doesn’t turn into tragedy.

Napster is a case in point. For a brief moment, I could get any song, any time, for free, and it only took a few minutes to download. That only lasted a few years. I recently bought an album on iTunes, and couldn’t share the song. There simply was no way to do it, and there used to be a way. This is disappointing, and illustrates the poignancy of a technology having its ‘moment in the sun’.

Given what you learnt, how should we as an industry think or act differently going forward?

Please, don’t worship innovation uncritically. Slow down. Stop to think about who and what is being cast aside. Consider your work in terms of justice and righteousness. Don’t sprint forward as fast as you can into the future, as this leads to technical and ethical errors that are impossible to get rid of.

We live in an innovation-centric culture. As a social critic I see the tension between innovation and the standards that enable us all to work together. Merely running head-first into the future creates long-term problems in exchange for short-term competitive advantage.

So a key role for technologists and entrepreneurs is to think about the potential unintended consequences of their work for society, and not just the benefits for them. Who might get harmed? Justice and righteousness are not something we train engineers and MBAs to do.

The other action I encourage people to take is to get to know how happenstance decisions in the past have limited your own understanding of possible futures. There is a horizon of possibilities we see in the future. Conventional wisdom and traditional assumptions exist in a narrow window of possibilities.

It is so easy to get caught up in conventional thinking. Step outside of this narrow view. I push my students in engineering and computer science to consider the value-sensitive design approach, pioneered by Batya Friedman and Helen Nissenbaum (among others).

Imagine we live in a world where the power of government to regulate is absent, and where engineers alone must be responsible for the things they create. This can inspire a new design ethos. Then imagine a world where engineers would find good reasons to work for government, and regulators would find enough resources to provide expert oversight. I get frustrated when students and professionals alike fail to imagine simple things like this.

I don’t want to come over as a ‘wet blanket’, as too often historians and philosophers can be seen as naggers and naysayers. I do want technology to work better. I don’t want to stop people building things. It’s OK to make money, too.

But what I really want is an Internet we can be proud of on a social level. When people in the future look back at us, let’s give them something to admire.

For the latest fresh thinking on telecommunications, please sign up for the free Geddes newsletter.