No, not you. Nor all the little children playing hide-and-seek (for their sakes, I wish them lots of shade). Nor do I refer to the song by a very young Frank Sinatra in 1994, an episode of the British TV series ‘Thriller’ aired on June 29th 1974, the orginal 1939 version of the Wizard of Oz featuring Judy Garland, or when Robert deNiro as Max Cady used these words in the movie Cape Fear(1991). Or any of the other 281 interesting entries on QuoDB which I have not checked yet. But you can see what the problem is, I suppose. I am at it again. The gathering of knowledge. Putting it all in a model. Having it for breakfast.

Felicitous holiday making

I have buried myself in the production of papers for my philosophy course for the past few weeks. Which is now finished! Everything has been checked, double checked, read by Husband, checked again – and handed in. Which is a weird feeling, suddenly having nothing to do. Well, other than the normal things. But no special brainfood. Husband and I had planned to take a short holiday to celebrate completion of the Grand 1/4 ReMa Philosophy – but the weather forecasts showed temperatures over 30, even over 35, so Husband asked: “did you take you free cancellation?” I get very grumpy when it is hot, you see. I had taken out free cancellation, so as we have air conditioning in the bedroom, which the holiday-home did not, I excercised the right to stay at home and swelter in peace. I have not made up my mind whether I am going to be working from home. Have done some calls and mail, but it might be nice to take the time off anyway. We will see.

I won’t be posting the finished papers as yet. Mainly because I am too nervous about them and want to wait for the results. I know, it is silly. I worked very hard at them and cannot imagine not getting a pass mark, but these philosophers are a funny lot. Remember my own grumpy-as-you-please professor complaining I had not distanced myself, in my draft paper, from some major debate which I had never heard of? As we say in Dutch, an accident may hiding in a small corner:

Actually, I finished my papers well in time, so I have been thinking since last week how to spend the holiday period in a felicitous manner. Do you like the word felicitous? It is standard jargon in philosophy of language, meaning that whatever is said is linguistically appropriate. J. L. Austin invented it, so that we can say of speech-acts if they make sense or not, rather than being true or false. I am proposing that the holiday-act is likewise neither true not false, but may be deemed felicitous if a) it is a break from work and b) I enjoy it and c) takes a considerable amount of time so I would not normally have time to do it. Right? Right. Let’s turn to my holiday project.

A nice little outline

I have been complaining on and off about how difficult it is to get a bird-eye view of ‘my’ area of philosophy, which is about language, mind, cognition and the evolution of our abilities in that area. So I thought I would do myself a nice little outline, starting with a text book by my philisophy-of-mind professor which should tell me what has happened in the last 30-odd years. The first chapter told me that there had been an paradigm-shift. Maybe you have heard of mind-body dualism, this idea that mind and mind and body are distinct and separate? That began with Descartes. Nobody believes that these days, or at least not officially. So the received view is that mind and body are not sepearate. That is as far as the agreement goes. Beyong that, there are many different opinions.

One of the hotly debated issues is currently about whether our thoughts are represented in our brain at a physcial level (representationalism). The picture below shows how that is supposed to work: when you see and egg, the image of the egg is represented in the brain. The opposing view, called embodied cognition, is that cogniton works in a totally different way, without representatons, but by shaping through the body and its interactions.

So there was me, happily charting the arguments against Descartes, when it suddenly occurred to me that my chart was not in fact about these philsosophers, or their theories, but really about statements which they did or did not agree on. You can see the problem below. I used the Rationale website to visualise the argument. The idea is that you collect argument for (green) and against red). The problem is the text inside the boxes. Sometimes that text represents a new theory, which is fine, but more often than not, these are statements. These statements will not stick to one line of argument or to one philosopher, but will pop up again and again as the building blocks of theories. In this case, I have modelled the arugment up to the point where two lines of arguments needed the same building block – which is something the rationale software cannot do, and I know from experience, messes up any mind map. Because mind-maps are not standardised. That is their beauty and greatest pitfall. So what to do?

There are actually two answers to that question. One is, if you do not know what to do, do nothing. Better to wait until you do know rather than forcing a situation out of frustration. The other answer is: find out! I am sure I need not worry about frustrating philosophers or philosophy (they are in the business of making problems for themselves), let’s try to find out. I have been grappling with the problem of systematizing knowledge all through my professional career-that-is-not-philosophy-but-information-technology, and I have never seen big solutions. However, I have seen a great many very expensive projects fail, so I will try not to do that. Looking afresh upon philosophy with my information-technology background, delivers the following insight: philsophy has not been modelled. At all. Everything is in books and papers. No indexes, no logical schemes, no roadmaps, no modelling tools – nothing at all. Just text. There is something called the Philospher’s Index which sounded very promising to my yearning ears. Unfortunately it is a bibliographical index. You can find any philsophical paper of interest there, and the editors have conveniently written an abstract and attached keywords for everyone of them. Nice, but not what I am looking for.

The “can” test

Time to engage in what Husband calls the “can” test. He claims that he has never met an IT person who passed it, apart from myself, but that must have been a fluke, because if I had thought about it, I would have given the wrong answer. Don’t tell Husband, it was only after the can-test that he proposed to me 🙂

Imagine you are somehow transported to another world. You are quite safe, but hungry. In front of you is a can of food (you know because there is a picture of stew or beans or whatever on the outside). How will you open the can? At this point all the IT people go off in search of a stone to hammer the can. But you are supposed to think: hey, this is a planet where they have cans, so therefore they will also have can openers. So you will go in search of a can opener or a person who has one. The test (I now know) derives from a old joke which was meant to mock economists, first mentioned in a book called “Economics as a Science” by Boulding. This is the joke:

A physicist, a chemist and an economist are stranded on a desert island and all they have to eat are cans of food. They’re discussing the best way to open them. The physicist says, let’s not overthink this – just bash them open with rocks! The chemist says, “No, we need to create a fire anyway and we can simply use the heat to cause the cans to burst open by themselves, and the food will already be cooked!” The economist thinks for a second and says, “First, let’s assume we have a can opener.”

https://en.wikipedia.org/wiki/Assume_a_can_opener

A stroll around the web – showed me that I was not alone in my queste for a can opener. I came across this fantastic website by a guy called Deniz Cem Önduygu who has charted the history of philosophy. I looked him up. He has an Masters in Visual art & communication. What a great profession! As for this effort: he calls it a “never-ending work in progress”. His major sources are “Bryan Magee’s The Story of Philosophy and Thomas Baldwin’s Contemporary Philosophy, but ideas of others and his own had been added. It is not complete, but only last year the site received a huge update. To give you a flavour of what he has done, look see, I have done you a selection with Wittgenstein. As you can see, a philosopher is associated with a set of statements; the green and the red lines go out to other philosophers meaning they agree or not. Every line can be followed. You can even put in search words. Enjoy it yourself, go over his site and play around. Be sure to switch off your VPN if you use one, because otherwise the graphics might not load.

Obviously this is much more than a nice little outline. The only drawback is, strangely enough, its contents: it is difficult to formulate a philosophical statement such that the original philosopher would agree with it. It is all about interpretation and understanding, and moreover, quite a few of them are dead now. The ones that live sometimes change their minds (as do the ones that have died, Wittgenstein for one). Still, if I knew how to do this, I would happily have tried my hand at creating my own version. Unfortutenatly the basic work is all done in excel and then professionally programmed (so the site says), so that is of no use to me. What is next?

The return of an old friend

If you want to create a representation of knowlegde that is not free-format, one way or another you end up at the semantic web. It is a web for data, and extension of the internet which can be processed by (mostly) machines. The semantic web has its own language, called Resource Description Framework (RDF). Basically, it is a data model, for which several syntaxes (serialisation formats) exist, such as N-triples, JSON-LD, Turtle, etc. On top of RDF other languages haven been built, such as OWL and RFDS. Now these names may seem just random collections of letters to you, and indeed, so they appeared to me. But some years ago I decided that the world really needed a good security taxonomy, so I dived right in.

To be precise, I dived into OWL, because this appeared to be the favourite language amongst semantic scholars. I installed the free OWL environment called Protege from Stanford University and worked my way through the “pizza example”. I came as far as a basic understanding how to construct an ontology (= a formal taxonomy written in OWL or another RDF based language) plus being to read one, but that was it. My problem was that to make a useful ontology you need to define many things in a specific technical or logical way. There was no gentle explanation beyond the pizza examples. I even posted a help-request on LinkedIn. Nothing. So in the end, after many many hours, I gave up. Too difficult for me. With friends like OWL you need no enemies, I thought.

Imagine my suprise at myself when I found myself back at the Protege tutorial, trying to set up a really simple ontology for what philosophers agree and disagree on, inspired by Deniz Cem Önduygu. I was not aiming for his wonderful red and green lines, but I was hoping to create something visual, say with bubbles. I did get a bit further this time. I succeeded in creating part of the ontology, but got stuck when I want to created relations between the various parts. The main thing was that now I understood that creating the ontology is only the first step. The next step is to put in the data, that is populate the ontology with philosophers, statements etc. And then, to create a program or some kind of view which allows you to use the insights which are part of the ontology. I imagined that I might succeed at step 1 and 2 if I spent another couple of days. But the last step was beyond me. I am not a programmer anymore. Or rather, the modern world no longer need an RPG-II programmer (not even a really good one) and I don’t know any modern programming languages. I suppose that learning Java or something like that just for this purpose would be a bit silly – once you stop using it, you forget it very quickly. So, yet again, I gave up. With a slightly better idea of what the problem is that I am trying to solve.

Another old friend

Before I returned to the field of security in my information-technology life, I spent some years on methods and techniques for application development. One of the issues was software quality, how to set and monitor requirements. I shall not bore you with the details, but it turns out that software quality has several aspects which partly depend on each other, and work out differently in differently programming languages and environment. To grasp the necessary characteristics, you need a different kind of database, one which does not force you to know beforehand how many levels of relations there will be between the top-level requirement and the bottom-level actions and measures. I learned this from a book by a researcher called Stefan Wagner. To accompany his book, he also developed a prototype, for which he used a graph database. This is a database where you store the relationship between data alongside the data themselves, as opposed to classic databases which have a predefined structure for storing data. With a graph database you can not only model but also graphically present data. Exactly, in bubbles. Just what I like. Unfortunately we did not use graph databases at the office. The programmer on my who tried to get Wagner’s prototype to work on a private computer, got hopelessly lost, but refused to admit it, thereby destroying the project. I am still a bit sad about that. Anyway, I can install whatever I like on my home computer, so I went ahead and got myself a community version of NEO4J. It turns out that you have to learn a query language called Cypher to be able to do anything at all. It is said that it is not difficult, but I really did not know where to start. I like to learn by example or trial-and-error. In this case, I did not even have a set of data to play around with. There was no help to be found anywhere. There are no manuals for straying philosophers with only a conceptual working knowledge of databases and query languages. So yet again, I gave up. At this point, Husband asked what the problem was. I had been swearing at my inability to understand something which herds of people were apparently very proficient at. He concluded that “it must be a different logic”. Hmm. Maybe. So I put NEO4J in with the other friend that I don’t need: Protege/Owl.

Between me and my data

By this point it was dawning on me that I was trying to solve a problem with another problem and so on. Yet another opportunity for the “can” test I though, perhaps someone else is getting as frustrated as I am about the same thing. You guessed right. People are. I found this wonderful visualisation of my problem. You won’t understand it, and neither do I, but my experimenting has taught me all the words that I do not want to be involved with. Basically, what I want is to achieve bliss, i.e. what happens at the very bottom (my insights), by connecting to a data-source somewhere on the web (the very top layer), without all the hassle of the in-between stuff. Simple, eh?

Notice there is some kind of magic connection called SPARQL which seems to do all I need? It turns out that SPARQL is not just a query language but that there are lots of SPARQL “endpoints” which you can connect to if you want data. What endpoint? Well, how about the whole of wikipedia? Or streetlevel crimes in Wales? Art in Holland? It is a motley collection to be sure, but growing steadily. You can even get an extension on your browser so you can check if the website you are browsing is using a linked-data format which would allow it to become a SPARQL endpoint.

Below the surface: structure!

If you have made it up to this point, chapeau! I tried not to scare you off with difficult words and ideas, but basically this is the story of how I got lost and found in this tangle, so I know how you feel. What is so great about this new discovery (which ofcourse is not so new, people have been working on it for years and years), is that there is structure below the surface of the internet. Have a look at this wikipedia page, on Wittgenstein. After you have done so, look at the image below. It is the very same wikipedia page, but seen through the “eyes” of the linked-data browser, showing all the RDF entries. Every wikipedia page is structured like this, which is why you can query Wikipedia from the outside!

Getting there in the end

So where does this take me? Well, a very long way. I have found several places on the web which will allow me to do a layman’s SPARQL query. For instance this one, called Sparklis. I can just ask questions like this:

And it gives me a list of exactly 17 philsophers that have pragmatism as a main field of interest. Or anything else I want to know! I can export these data, or reuse the query in another program, which I have also been playing with, called Gephi, which will construct graphs for me. Yes, my bubbles. Finally I am getting to the kind of questions that I wanted to anser for my felicitous 🙂 holiday hobby project. I want to be able to tell how a philsopher was influenced by other philosophers, so I can decide on which common link to read up on. Or not. Saves quite some searching around, and the feeling of being able to see, in a virtual sense, the big picture, and draw conclusions from it, is great. I love it. Have a look at some examples what I have found so far. I will put them in with the concept side of my blog.