Konstantin Tovstiadi

There are many ways to learn about a new city or a country. You could look at its GDP data, or the exchange rate, or the atmospheric pressure and dew point, or any other point estimates and indicators. You could read a Lonely Planet guide about it. Or you could go spend a week there. If somebody asked you about it later and all you had was the GDP data, you could simply tell them what the number was. They would have the same information as you then. But you can't share your travels with them the same way. Your impressions are complex, multifaceted, irreducible to stories and pictures.

Complex phenomena are all around us - snowflakes, tornadoes, ant colonies, stock markets. Studying them requires a new approach. Converting them to point estimates - size of the flake, speed of the wind inside the tornado, number of ants, Dow Jones - doesn't give them justice. The simplifications are too crude, too much data is lost. One solution to that is to have multiple point estimates. We can map the snow flake configuration with 1 micron accuracy - and if that is not enough, even greater detail.

But it is never going to be enough. The map can never be the territory. The measurements are imprecise; initial imprecision translates into greater and greater errors as we model the reality over time using point estimates. Eventually we have to admit that some of these problems - even the seemingly simple ones - are transcomputational - we cannot know what the weather will be two weeks from now, what the stock market will do in a month, or what would be the in-thing at the Milan fashion show next year.

Stuart Kauffman tackles the issue in his new book, Reinventing the sacred. Kauffman is one of the most respected scholars in the study of complex systems. This is his fourth book, and it builds on the first three. His earliest book, Origins of order, published in 1993, is a rigorous academic account of order-for-free, bottom-up emergence of high level order from the actions of lower level agents - such as the water molecules forming a snowflake. It was followed by At home in the universe, a brilliant translation of the first book into language comprehensible to the non-academic audiences. In 2000, Kauffman published Investigations - although it would be more accurate to call it Musings or Ramblings. It reads like a notebook - powerful raw ideas in need of better argumentation and cleaner style.

Reinventing the sacred resembles all three. It has some hardcore theory that many will hard to follow (such as the discussion of quantum decoherence). It has some excellent explanations of what these concepts mean in layman terms. And it has oodles of speculation and conjecture.

But there is also a new element - one that has to do with religion, ethics, philosophy, and the emergence of a global society and culture. Emergent order is the quality of the universe. Life emerges naturally. Human consciousness emerges naturally. Culture, language, society, economies all follow similar paths - unpredictable, irreducible, forever changing, forever creating. There is no need to invent a creator God to explain it all. But there is something that can be revered - the endless creativity of the universe itself. Worshipping this creativity is the suggested foundation for a new global culture, global ethic, global religion.

It is a beautiful idea, but one that is not easy to argue. The argument annoys the hard scientists (and anyone who wants to defend reductionism), philosophers (and anyone who finds Kauffman's foray into epistemology amateurish), complexity scientists (why can't Kauffman talk about biology and leave the rest out), and last but not least, the believers in a Creator God.

The task is enormous, and it is also enormously important. Kauffman's book isn't flawless - but on a subject as daunting and broad as world ethics, no argument can be. The book suffers from occasional repetitiveness and uneven style. It oscillates between barely understandable discussions of quantum physics and bar-style down-to-earth real life examples delivered with excessive nonchalance and camaraderie. But most of the time the book stays right in the middle, and despite redundancies and style issues, it delivers a strong argument - an argument that is not only thought-provoking, but timely and important as well. I have thoroughly enjoyed it - how can you dislike a book where a scientist who understands biology, mathematics, and quantum physics devotes several pages to praising rhetoric?

But this is a reductionist narrative about the book. The book is a lot more complex than my ramblings about it. You should read it yourself, or at the very least, listen to Stuart Kauffman talk about it himself. You can also read some of his most recent thoughts on this NPR blog.

In the 1990s, all books in intercultural communication started with "we have never seen so much international travel / globalization". In the 2000s, all computer mediated communication books went on about never-before-seen ways to interact online. And in the first week of the 2010s (that The Economist suggests we call "teens", "decadents" or "debtcade") I managed to screw up in an entirely new, novel, never-before-possible way.

I have been talking about the software stack; and how depending on your geekiness you either live at the very top of it or you can actually penetrate its murky depths. The stack situation, in the great scheme of things, is a consequence of the division of labor and specialization. Even the geekiest nerds don't know every element of the stack. Generally, any modern household, vehicle, appliance, is a stack where the user knows only the top layer. It used to be that a peasant could fix the tools of his trade himself. There was nothing mysterious about a sickle. A John Deere tractor is a different story.

So armed with a minimal knowledge of server technology, I spent the last week of the "oughts" setting up a Mac OS X server - all the way from network connections and DNS records to remote administration, databases, and active PHP applications. And then I managed to lock myself out of one of my accounts; and decided to re-install everything.

It was all going well in the beginning - faster than before, because now I knew what I was doing. But it all went to hell in a hand basket when it was time to set up the database server. You start the MySQL daemon with a click of a mouse; but then you have to use a third party program like Navicat to create users and databases. The first time I was doing the server setup, I looked it up online, found an answer that worked, and was creating databases in five minutes. But the second time was different. I still didn't know how to set up the database server; but I knew more about the whole process - enough to be doing the Google search for instructions with different terms, enough to not be able to find the solution that had worked before. So knowing more about the process was actually counterproductive.

In 2000, I wanted my own web page. Getting one felt like magic. Then I wanted my own site. Another magical experience. Then my own virtual server. More magic. Then a chunk of a real server that somebody else managed. Kind of magical. Finally, a server all to myself to administer. Today is finally up and running, in all its 16-virtual-cores, 2-terabytes-of-online-storage, 90-gigaflops-of-computing-power glory.

So I am looking for an open-source CMS to run on it; and given Drupal's recent runaway success with kings and presidents alike, Drupal is my top choice. So I am reading e-books about Drupal; and I come across a description of Drupal's architecture and its place in the "software stack".

The Drupal code is written in PHP, a programming language designed to deliver dynamic content (rather than static HTML). The application relies on the data in the stack layer below it - the database layer, in Drupal's case MySQL. Deeper still below that is the web server layer, enabling both PHP and MySQL function over the Internet (in our case, the web server is running on Apache). At the very bottom, the web server is embedded in the operating system (we use Mac OS X Server, but it could be Windows, UNIX, or Linux). So all in all, four layers:

PHP > MySQL > Apache > Mac OS X

You could extend the stack both up and down. The content management administrator doesn't interact directly with the PHP code; he or she normally uses a graphic user interface (GUI) to access the application. The administrator sets up the experience for the end users, who interact with the system through their own GUI. On the other end, for the operating system to run, the server hardware must have some basic software (=firmware) that allows to install the operating system, and check the integrity of the hardware. So the extended stack is more like this:

End user GUI > Administrator GUI > Application > Database > Operating System > Firmware

The IT staff's goal is to make as much of the stack invisible to the end user. The goal of the advanced user is to understand how the stack functions as far down as possible. The deeper you can reach, the more power you have.

But the beauty of today's situation is that one can have power even with just top-of-the-stack access. Modern servers come ready with the whole stack set up and ready to go. There is open source software for all the layers of the stack. Besides the depth, there is breadth - there are open source applications for almost everything under the sun. Which means the end user no longer needs to dive into the stack and get lost in heady geeky things. They can stay on top of the stack and focus on the task no one else can do: creating and organizing content.

Thirteen years ago I made my first trip abroad - three weeks at a language school in Brighton, UK. I had to obtain a British visa. It took two interviews and two visa officers to grant a visa. The second officer started the interview with a let's-get-to-the-heart-of-the-matter question: "Why did you give our visa officer a cookery book?"

It wasn't a cookery book. It was the project that I was working on - relating culinary traditions and culture, showing food as an integral part of history, culture, and ultimately language use. What better way to savor them than through food?

I was happy to find out that my two favorite things from Portugal - fado music and port wine - are related, too. Let me tell you how.

Fado is at least a couple of centuries old. It has its roots in Afro-Brazilian music (Brazil being a Portuguese colony at the time). It means fate or destiny in Portuguese. Fado develops in Lisbon and Coimbra in bars, taverns, and brothels - a folk tradition yet a distinctly urban one. Almost always performed in a minor key, it is a melancholic song with profound lyrics about life, love, and death.

Fado singers can be accompanied by various instruments, but almost always by two types of guitar - the familiar Spanish six string version (against the common wisdom of Northern Portugal that only bad wind and bad marriages come from Spain*) and then a more exotic twelve string Portuguese guitar. This instrument is a variation on the English lute - it comes to portugal via Porto, the port city at the mouth of river Douro, a trade center. Porto the city is known best for its eponymous staple - the port wine. The port wine growing region (a time-honored tradition, demarcated a century before Bordeaux) is the upper basin of Douro, in the mountains near the Spanish border. For centuries, Britain was the chief consumer of port wine; British merchants took great care to control not only wine trade, but wine production as well. The names of the major port houses bear witness to it to this day - Graham, Warre, Dow, Taylor Fladgate, Churchill, Sandeman, Smith Woodhouse.

Port deserves its own story - at least one, possibly many - best told by the port makers, such as the Symington family. Let's get back fado. There are many phenomenal fado singers, past and current; my favorite is Mariza. Her 2008 concert in Lisbon is superb. You can find it on iTunes, and have instant gratification; but it is best to have patience and get the CD version because it comes with a second disc - a DVD with a documentary called "Mariza and the story of fado".

The second visa officer understood the fun of relating food and culture. Before long, I had her laughing and smiling. When the ice was broken, I knew for sure my visa application would not be rejected.

*De Espanha, nem bom vento, nem bom casamento

If the audiences are everything for web projects, then a successful web site is one that finds a way to engage its audience - i.e. finds ways to grow through reader contributions, comments, ratings, and usage patterns. The same applies to other projects that involve collaboration - especially to the development of online collaborative tools for academics, such as institutional repositories (IRs). The authors of the key text on IRs describe the struggles of engaging the audience very well:

Acquiring the content is slow and laborious work, and at the present time we pay for it with the sweat of our brow, rather than by dipping into our materials budget. It involves - in addition to the seminars in the library - working through lists of academics with research management responsibilities, research journal editors and senior managers, as well as 'common or garden' academics. It requires the organization and tenacity (and thick skin) to lobby powerful committees, to meet academics in their own territory - in departmental meetings, and at lunchtime seminars held by research groups. It requires the hunting down of unorganized self-archived work, and the corralling of it within the institution's managed repository. It requires the capture of author permissions and an awareness of publisher policies on copyright transfer. It involves knowledge of licensing as it applies to self-archived and open access content. ANd most of all, perhaps, it involves the practice of constant repetition of the same message over and over many thousands of times, often to the same people, because the issues are difficult and non-intuitive to academic authors. This is a liaison work, and in a large research university it can easily cost a full-time professional member of staff. (pp. 38-39).

Jones, Richard; Andrew, Theo, and MacColl, John. (2006). The institutional repository. Oxford: Chandos.

I made my first web site in 2000. I can now claim I have been working with web technology for a decade. And since my last three jobs are directly related to delivering web content, I feel like I have a grasp on what it takes to create and run a successful web project.

I have talked about it here before, focusing on three major concepts: form, content, and audience. I have been rethinking the relationship between the three lately, and here's an updated version.

Form is really synonymous with design. These three concepts - form, content, audience - are about what you deliver (content), to whom (audience), and how you present it (form / design).

I have argued against a workflow that starts with design, then moves on to the content, and then tries to adjust to the needs of the audience. I have claimed that form is nothing, content is nothing, audience is everything. I would add one more thing today.

A good web workflow works in the reverse order. It starts with an audience analysis - what do they need? Once the needs are clear, what content can be created (by us, or better still by us with the audience's feedback or contribution)? And once we know the content, how do we design the interface to deliver the content in the most accessible and intuitive way possible?

Technology is part of every step of this process, but it should never ever be a deciding factor driving the change. You don't update the site just because it is time. Or worse still, because there is some new flash(y) technology available. You use the technology that allows you create the good interface you need to deliver the content that your audience wants.

It starts with the audience, never with technology.

When presented with a problem, you have to find a solution. Sometimes it is straightforward. You walk up to a store, the door says "push", you push and get in. Other times it is a little trickier - there is no sign, but you can observe the people coming and going and note which way the door is flapping and apply the observations to get in. Still other times it is trickier still - there are no signs and no people - and maybe three different doors. Which one do you pick, and do you push it or pull it? Do you enter quickly and recklessly - or carefully pausing to close the door as silently as possible?

The tricky problems are usually the ones requiring cultural knowledge. The locals and the insiders have the knowledge, you don't. Feeling like an outsider sucks; besides it costs money. The locals know all the tricks: where to make a short cut, where to park for free, where to buy discounted rail passes.

When you are a cultural outsider in a new country or a city as large and complex as Moscow (where you don't visit for two years and it changes enough to be confusing), you are constantly presented with these tricky cultural choices. And unless you stay for a while, you only get a single trial on solving each one of them correctly.

That means you will make mistakes and make inefficient choices. You will open the wrong door, take the long way around, and buy expensive tickets. It will cost more than it would cost a local. You might even get into some embarrassing situations. But you just have to let go. Enjoy the few that you get right, and don't judge yourself for the ones you don't. Enjoy the new ride and ignore the bumps. One day you will be experienced enough to get almost everything right on a single attempt, and composed enough to meet the few occasional failures with a smile.