Co-Immunology and the Web

For a Healthier World

Henry Story
20 min readMar 29, 2020
Dali’s attempt in 1936 to read ”Authentic Paranoid Fantasies” in a diving suit.

The COVID-19 global pandemic has taught us that immunology is not just a biological phenomenon but a personal one requiring us to adopt hygienic practices, a social one, and indeed with the closing of frontiers a geopolitical one. But it also breaks into the space of technology with requirements that we learn hygiene rules about reposting content to limit the transmission of false information about the virus.
I will show how the concept of co-immunity developed by Peter Sloterdijk can give us a framework to understand this. The essay ends by developing an analysis of the immunological features of the Internet and locating pieces that we need to make a healthier World Wide Web.

What we have learned

We have all learned about this virus — a measly strand of genetic code — can only replicate by using a host cell. We have seen detailed reports on how the immune system functions to fight the virus by first slowing down the functioning of the body — leading to a rise in temperature — and then locating and destroying infected cells. (see the informative World Health Organisation video “The Coronavirus Explained and what you should do”)

WHO video “The Coronavirus explained and what you should do

We have seen how the virus is transmitted through coughing and touching, and that individually we can fight it by washing our hands or wearing face masks. We also saw how we need to work together by physical distancing, isolating ourselves and reducing physical contact. As social beings, we depend on others for food, heat and psychological well being: we thus need to protect ourselves in order to protect others, and so to protect ourselves in turn. We thus virtualised our social interactions by teleconferencing and working from home. In short, we saw how immunity requires co-immunity.

We can wear masks to protect ourselves, but this can, of course, be taken too far. Sloterdijk illustrates this in his third volume of Spheres, by describing the happening that Salvador Dalí undertook in his 1936 London exhibition on the subconscious, which he recalled as follows in his memoirs:

I had determined to give a talk during the exhibition in a diving suite as a representation of the subconscious. I was placed in the armor and even fitted with heavy lead shoes, making it completely impossible for me to move my legs. I had to be carried onto the stage. Then they placed the helmet on my head and screwed it tightly shut. I started my talk behind the thick glass in front of a microphone that clearly could not record anything. Nevertheless my mimicry fascinated the public. Soon though I ran out of air, my face turned red, then blue, and my eyes turned upwards. Clearly they had forgotten to provide me with access to air, and I was close to asphyxiation. I signaled to my friends through desperate gestures that my situation was becoming critical. One of them ran for some scissors and tried without success to puncture the costume; another one tried to unscrew the helmet, and as that did not succeed he started banging on the screws with a hammer… Two men tried to tear off the helmet and a third one continued whacking the metal so hard that I nearly lost consciousness.
On the stage one could only see a mass of wildly moving hands, from which I emerged now and again like a dismembered puppet, and my helmet sounded like a gong. The public applauded heavily to this successful Daliesque melodrama, which in their eyes clearly represented how consciousness was trying to communicate with the unconscious. I though nearly died during this triumph. As they finally ripped off my helmet I was as white as Jesus as he returned after his forty days of fasting from the desert.

A virus is a tiny piece of harmful information that travels in the air we breathe. There are many other such pieces of genetic code in the air, not all of which are dangerous, indeed many are helpful. The paranoid fantasy is to think we could disconnect ourselves from the world and connect ourselves to a sterilised air supply to be totally autonomous. But then we’d have trouble communicating (as he did) and be dependent on the technical apparatus functioning (which in his case it did not). In any case, we remain dependent on others to come to our rescue if it stops working. So we stay co-dependent.

We depend on others at the social immunological level. On doctors, and nurses first of all, on scientists to analyse the virus, on people working in shops to provide food, on electrical and internet companies to continue working, on technologists to build ventilators and testing equipment, and sadly on police to enforce distancing restrictions.

The better the testing equipment, the easier it will be to fight the virus intelligently, by quarantining as few people as possible and allowing the economy to get back on its feet. South Korea, for example, has produced booths to test people in 7 minutes according to this video.

Indeed the idea of the body politic, so well represented by Hobbes on the front cover of the Leviathan, has taken on real meaning with recent events. See the recent Guardian illustration below.

We have seen nations close borders and impose new borders internally. Nations that just a year ago were critical of the UK for voting for Brexit. We have seen international flights slowed down then closed, as governments take distancing measures for the same reasons as we do: to stem the spread of the virus. Nations, too, work on a co-immunological basis then.

The concept of co-immunology

The concept of co-immunology is not from me but from Peter Sloterdijk, one of the most famous living German philosophers. In his trilogy Spheres (published by MIT Press) he retells the history of civilisation from an immunological perspective. That is very good if long, read. Luckily Rolf Rauschenbach published a 1 summary of Sloterdijk’s work in 2011 “How to govern the universalising community: Peter Sloterdijk’s concept of co-immunism” which concentrates on the aspects of the work that concerns us here.

Sloterdijk starts from the view that “Co-existence precedes existence”. We are always enveloped in spheres in contact with other spheres on which we co-depend. Starting from the fetus in the womb of the mother to the child learning from his family, then moving into schools, and integrating further institutions that form the body politic, which itself depends yet on other nations for commerce and know-how.

Each of these spheres of being needs to be able to distinguish the inside and the outside in order, first of all, to use external energy to keep order inside, fighting the unavoidable process of entropy, which the bible renders as “Ashes to Ashes and Dust to Dust”. This keeping of order was conceptualised in 1943 by Quantum Physicist Erwin Schrödinger in his lectures “What is Life?” in terms of negative entropy. Around the same time, negative entropy was equated by Claude Shannon with information.

These spheres also need to fight being repurposed to the aims of a different living being, with goals incompatible with those of the larger structure of which they are part. From a physical point of view, the virus is just another process. But from the point of view of the larger structure in which the cell finds itself, it is destructive. From the physical point of view, strings of differences that flow down the internet pipes that we think of as 0s and 1s abide by the same laws as the rest of the universe. But from the point of view of the body politic, some constitute disinformation leading people to adopt dangerous behaviours, and others constitute information that leads to curative behaviour.

We thus have three levels of immunities: biological, psychological and social. One should perhaps add hyper-social for the relations between states that have, over the last century, put in place institutions to reduce the calamities of war. Each of these depends on the other. The body’s immunity is improved by individuals learning to adopt hygienic practices; those are helped by quarantining policies at the local and international levels. At each level, communication is restricted but cannot be stopped: trade, especially for medical goods and food, must continue while new processes are put in place further to limit the spread of this microscopically small virus.

Rolf Rauschenbach then summarises Peter Sloterdijk’s spherical reconstruction of the evolution of human society, into a micro-spherical pre-historic period, a macro-spherical period of empires attempting larger and larger unities ending in the second world war, and a plural spherical one, acknowledging the many interconnected dimensions of spheres with no centre of control, which will require a hyper-politics, a concept Sloterdijk does not develop much, but for which we can next give a technical grounding.

Co-Immunology and the Web

The internet is one of the largest technical systems to have evolved on earth. To succeed, it had to convince all of the relevant actors of its advantages. These actors included first of all the nation-states; the military complexes who needed this to communicate in case of the ever-present danger of nuclear war; the telecommunication companies mostly anchored in territorial networks; the transnational companies who built the infrastructure of cables crossing the oceans and satellites orbiting the globe; the business community who would gain in efficiency; the computer companies that would then be able to sell devices attractive and enticing enough to use, to create a market that would reduce the cost of the whole project. As a result, the internet architecture had to be peer to peer and the protocols free to allow all these actors to participate without the danger of having to pay a tax to a gatekeeper. Such an actor would have posed an unacceptable threat to their autonomy, also known as sovereignty.

In contrast, the Minitel hyper-text system deployed in the early 1980ies into every French home was controlled and owned by the French state. It did not, therefore, satisfy the above requirements concerning sovereignty and so could not scale to the globe.

The Web of Hyper-Text Documents accessible from any Browser has a P2P architecture.

By creating an open protocol and document format that anyone could implement and participate in, Tim Berners-Lee developed a decentralised evolving hypertext platform that added the killer app to the internet: The World Wide Web. Anyone with enough skills could build a web server or a browser. Much more importantly, once those were available, it needed only minimal skills for someone to publish web pages using the HTML format and to link their page to other pages they trusted. This light-weight peer to peer linking and weaving, built on trusting others to continue publishing the same content at the same place as well as to link to other trustworthy content, created what could be called the society of the Web. Co-immunologically speaking, the Web built on people developing hygiene skills to link only to quality pages that would increase the value of linking to their web page. The suspicious pages would not get linked to, other than perhaps by other sites of the same ilk, and these were quite easily recognisable. The most important currency at the time was trust. It is on this widely deployed trust that Google could deploy its PageRank algorithm, allowing them to bubble up quality content as per magic to the top. The quality of such pages could then easily be increased by Google identifying pages published by institutions of knowledge such as universities, government departments or the health care sector.

Initially, the Web had no security infrastructure. Documents were sent in the clear over unsecured connections, which would have made it easy to create a man in the middle attack. This was actually considered an advantage: the internet was slow and so national web caches had a significant role to play in speeding up communication. If one user in England fetched a page located in a server in the US, the UK cache would keep a copy of this page to allow any other user to fetch it later without requiring the packets to cross the Atlantic. This did not pose much of a problem as the value of the Web as a project was not clear enough for anyone to see much value in subverting it. But it also meant that there were official intermediaries potentially able to oversee content as it flowed across the globe, creating a simple surveillance architecture limiting certain excesses.

This lack of security at the transmission level had to be remedied to allow commerce to flourish on the Web. With financial transactions on the line, it would have been too easy to capture credit cards flowing over the wire, allowing those who did this to live at the expense of others. Thus Netscape developed the Secure Sockets Layer SSL and deployed it after a few false starts in 1997, starting thereby the dot-com boom. This led to their emergence as a giant on the stock market, soon followed by others such as eBay and Amazon that we still know today. It also made it possible for web sites to identify themselves cryptographically to users with X509 Certificates, increasing people’s trust that they were purchasing good from the right website.
Co-immunologically: the addition of commerce changed the game of trust on the Web to make it valuable enough to subvert link-based trust, requiring the communication protocols to be secured. It led to the creation of private institutions of trust known as Certificate Authorities, leading to vast fortunes such as that of Mark Shuttleworth who sold his company Thawte Consulting for a value of around 600 million dollars in 1999. (He then reinvested the money in open source software distribution company Ubuntu)

The blogosphere’s meteoric rise from 1999 onward pursued the philosophy of link-based trust that started the Web. By helping authors publish content and allowing people to subscribe to their changes, it became possible for an author to create a reputation online and allow readers to follow the latest updates without having to wait for months before the search engines web crawlers came across the content again.

Yet the growth stalled around 2007. Facebook had emerged and enabled two things the blogosphere had not dealt with properly: social network-based access control on content and live comments. The ability to have a discussion around content with readers, who could point out flaws or bring new ideas to a text, made blogging into a completely new medium of research. But it also gave incentives for others to create comment-producing robots with the sole intention of adding links to websites containing advertising, increasing Google Page Rank positioning or link to pages with phishing scams. As a result, comments had to be reviewed at the author’s leisure, making it difficult to have live conversations on a topic.

The Application Web developed siloed data architecture with Apps designed for a particular domain.

Co-immunologically: the blogosphere strengthened various strands of link-based viri who would feed off the advertising industry or phishing scams. The only means of protection against it required slowing down the blogging process, leading to the Macro-Spherical growth of Facebook and Twitter concentrating all the data in one provider.
These Macro Spherical projects, requiring all discussions to occur on the same platform, were themselves fed by advertising-based revenue (even if initially they denied it). We can think of advertising as a type of information virus that travels along the usual channels of communication. Many of these are harmless or beneficial, but some are not. The question is: beneficial to whom? These Macro-Spherical projects saw all these viruses as benefitting them as they provided food for their growth. As a result, they did not care to enhance the critical/hygienic know-how of their users, which could have stemmed the duplication of dangerous content. Instead, they made duplication as easy as possible and furthermore made it easy for other agencies to pay for the spread of content. Where the initial ideal was to use people’s ability to filter content, their focus on growth led them to build one where people lost their critical barings.

The Macro-Spherical extension beyond its abilities is the problem all imperial projects face. The space they create leads to the growth of new types of viruses, which we now recognise in all kinds of guises such as fake news, misinformation or disinformation campaigns. But they then also give rise to counter-spheres such as the geopolitically driven one developed in China, Russia and elsewhere, described in The Four Internets.

Any hope to solve these problems by AI alone without using politically based institutions is bound to fail. AI can detect patterns in data, but it cannot evaluate the truth of these: without police on the ground and judicial systems evaluating each case on an individual basis, there can be no way for AI to distinguish truth from falsehood, or make just decisions. (See the online book Law For Computer Scientists which has a chapter on AI.)

These silos, however huge cannot, due to their architecture, address the genuine and much bigger needs for co-operation required by our hyper-industrialised societies, and so are themselves doomed to stall. As we saw earlier: societies are complex social networks linking citizens to each other, to institutions (health care, schools, universities, armies, police), private firms and global players of various sorts. Each of these need to both maintain their sovereignty as well as co-operate with others that have the same requirements. One needs immunological systems to be developed at all levels. It is not realistic to have police services and health care systems use Facebook or Twitter as their main platform. Apart from the fact that the data structures they make available do not address the needs of these services, in the completely unlikely case they did, the requirements on sovereignty would constitute an insurmountable barrier (see Epistemology in the Cloud).

To answer this we must design technical systems that must both respect the sovereignty of each actor and their need for cooperation. Each actor has to be able to control their data, yet also to cooperate with others. Each actor needs to be able to have its own immunological system, even as they depend co-immunologically on others.

The way to address this is to extend what has been shown to work. We can extend the hyper-text Web to a hyper-data web as has been shown by the Linked Open Data Community. But to speed up its evolution, we need to reduce the cost of publishing. This requires open protocols for reading and writing data (LDP). For security, we need non-centralised identity systems (WebID), extensible certificates (Verifiable Credentials) and decentralised access control systems. Most important of all, are that it requires applications equivalent to web browsers (HyperApps) to be written that elegantly display such hyperdata without being tied to a particular service. These apps must, of course, do the work of publishing the content and data themselves (hence LDP).

Hyper-Apps following links of hyper-data and displaying it to the user, reintroduce the lost P2P architecture.

This is what Tim Berners-Lee’s new Solid project is about. It uses hyper-data (a.k.a Linked Data) to allow any individual or institution to create a Web Identity, link her content to others using hyper-applications that can read and write to containers protected by Access Control Rules. These rules are themselves, of course, described using hyper-data. Instead of requiring medieval-type motes and walls to surround an organisation (intranets), each piece of content can have its own immunological protective barrier, that is nevertheless open to interaction with the outside world. This closed/open system works just the way biological systems do, allowing larger systems to be formed out of smaller co-operating ones, each one with various protective barriers, working together as a co-immunitary system.

This decentralised, flexible social network can work at all levels of Peter Solterdijck’s spherology:

  • at the microspherical family level, each house can have it’s Freedom Box installed, monitoring and controlling input and output of data including that from the ever-growing Internet of Things.
  • Companies and institutions can also run Solid based servers, allowing them to cooperate on various projects efficiently, making available in a few clicks the information needed by other companies in their supply chain, or in order to communicate with their clients, research institutions, law firms, or governments.
  • National Health Systems and doctors could make data available securely to each citizen using the same protocols. Research institutions, bound by the right privacy requirements, be given access easily to data when needed, enabling them to develop cures in case of emergency.
  • Banks and supermarkets could publish access-controlled data about their clients’ transactions enabling apps to be developed to help automate accounting tasks, and so giving people a better view of their financial health.
  • If one added micropayments then content producers such as musicians could publish their own music, that could be purchased using hyper-music apps.
  • Finally, Nations that are themselves constituted of institutions can publish machine-readable records of the data about those institutions and companies that make them up, as well as publish diplomatic links to other countries. I call this the Web of Nations (a technical implementation of the philosophical explorations by Bernard Stiegler on internations). It could enable Web Browsers to show rich data about the legal location of web sites users land on, telling them in which country the company behind them is based, what type of company it is, and the legal of financial health of the company. This is one of the 13 use cases developed in the article below. It shows how the geopolitical Web of Nations can be used to help increase the immunological systems at the lower levels.

To put this in Sloterdijk’s terms, this multiplicity (foam) of interconnected spheres based on hyper-links is what is needed to create a global informational immune system for a new hyper-political space.

Improving the Climate of Trust: Memory

A recurrent topic in Sloterijk’s Spheres is that of climate, atmosphere and atmotechnology. In our information space, the climate of Trust is essential. As we saw, the Web relies principally on trust networks of hyper-links. What can we do to improve trust in our atmotechnology?

Immunological systems are retentional ones. This is well known for the endosomatic layer: our immune system learns with each infection to recognise the intruder, allowing it to detect new attacks earlier and so reduce the damage. This is what makes vaccinations so effective. This can be extended to the exosomatic layers as argued by Daniel Ross: our various social immune systems of which the legal systems, use past cases transmitted in writing, taking the particularity of the current case into account, to come to a judgement. (We here rejoin Robert Brandom’s analysis of the law.)

If we think of hyper-links as the paradigmatic technical device implementing the transitional on the Web, then we can see that this is also where questions of co-immunity will appear most clearly. It is in this transitional space that work needs to be done to increase the trust publishers and people following links can have that the document linked to has not changed. (see §4.6 Persistence and Preservation of Sarven Capadisli’s thesis) This would help preserve the integrity of the Web, reduce further types of viral attacks that could arise by changes to the linked-to content and would create an atmosphere in which linking could flourish. Here are some technical paths to explore:

  1. Archives: Clients and servers should be able to keep copies of remote content so that if linked-to content disappears (e.g. when someone dies) those linking can remember the content they had linked to. If the content of the copy is clearly attributed along with the original metadata, as well as protected, this amounts to a reasonable right to not forgetting.
  2. Versioned URLs: These are URLs structured to contain a way to verify the content. This can make the Web more brittle. What happens if content changes? If any small change to the content opens up a huge warning sign, or makes access impossible then this would be unacceptable. But if instead any mismatch is used as a warning signal, this would allow the current state to be compared to a cached version, perhaps the version from the linking page.
  3. Versioned Data: Web Resources should point to previous versions so that histories of changes can be understood and followed.
  4. Notification of changes: protocols for notifying people linking to a document of a change should be developed and widely deployed.
  5. IPFS integration with Solid: The strategy of IPFS of allowing content to be duplicated everywhere may not be for all use cases. For example, it makes it difficult to change access control rules, and many actors prefer some data even if encrypted not to be widely distributed. But it does provide a way to deal with the previous 3 points.

Conclusion

The Corona Virus has made us aware of the many dimensions of immunity and co-immunity. Fighting it requires not just individual change of behaviour, but also social, political and geopolitical changes. We have seen various immune systems operating at each of these levels, and the need for improved technical ones. To fight a biological-based virus that is a genetically coded piece of duplicating information, the internet was used (along with the press, radio and TV) to inform people of the dangers it posed and the physical distancing measures that could help. Yet, these media themselves are host to their own forms of viri (fake news, disinformation campaigns, …) that act as if they were a technological extension of SARS-Cov-2. This has forced the major social networks to clamp down on such misinformation that until recently had found little reason to curb (see Twitter’s policy, for example).
How did they determine what constituted misinformation? For this, they had to rely on locally-based institutions of knowledge (universities, hospitals, health and legal systems, police). Just as executives in each country had to pore over such reports (e.g. the Imperial College Report. See also the detailed article from The New Statesman and the update at the end of this essay) in order to implement the right confinement measures and support R&D to reduce the long term cost.

This highly decentralised cooperation between institutions nationally and internationally can be improved by extending the Web to enable the secure exchange of data which is what the Solid project aims to do. By allowing each resource to be protected by, among other things, flexible access control rules we can move away from medieval firewall based security to a world where each sovereign actor can also easily cooperate with others using hyper-Applications. For this, we have to think not just in terms of immunity but also in terms of co-immunity.

Addendum

For a different overview of Peter Sloterdijk’s work that concentrates on the poetic creation of space, see Pieter Lemmen’s review.

For a deep analysis of the relation between Stiegler and Sloterdijk’s work starting from the Presocratics see Daniel Ross’s chapter for his soon to be published book “The End of the Metaphysics of Being and the Beginning of the Metacosmics of Entropy”.

In “Psychic and Collective Anaphylaxis: For an Organological Critique of Sovereignty”, Daniel Ross develops in depth the philosophical themes and tensions for which The Web of Nations proposal could provide a cure at the epistemological level. (He uses the word “critique” in the philosophical sense, which is not to be understood as “criticism” but as a process of deconstructing and reconstructing the concept to allow it to be renewed in the current maelstrom of technical change). He shows the importance of the process that legal systems based on the written law require, the reflective time this necessarily takes, and contrasts this to the speed at which the technosphere is evolving.
I believe that by making factual aspects of these laws machine-readable — such as company registration records — and placing those on the Web to be used by browsers, these legal hyper-data will help improve the epistemic grounding of citizens globally, without denying the territoriality of the law that Daniel Ross also thematises. If this is not done, then the problems he describes so well will only get worse.

On the Imperial College model that led to the change of strategy in the UK see:

📝 Save this story in Journal.

👩‍💻 Wake up every Sunday morning to the week’s most noteworthy stories in Tech waiting in your inbox. Read the Noteworthy in Tech newsletter.

--

--

Henry Story

is writing his PhD on http://co-operating.systems/ . A Social Web Architect, he develops in Scala ideas guided by Philosophy, and a little Category Theory.