Applying methods of data science to philosophy

[Firstly, I will admit that this post is part of the problem it diagnoses.] Recently watching Hans Rosling‘s rather fun “The Joy of Stats”, I encountered Microsoft Research‘s Head of Computational Science, Stephen Emmott, discussing how advances in statistics and computation are leading the way towards a new model of science. Where previously, he says, science worked according to experiment and hypothesis, our new ability to process vast amounts of data as never before is in fact opoening up new realms of study, allowing us to make new proposals and even to ask entirely new kinds of questions. We have changed the words, and now we are playing around with the syntax and grammar. (Link to Dr. Nico Sommerdijk of Eindhoven University of Technology discussing the same matter here)

Philosophy has only just recently begun to move beyond the old, even scholastic model of philosophizing, to make the possibility of “experimental philosophy” explicit (the main e.p. blog is here.) Now, I will emphasize here at the outset that we need to deal with the question of whether we regard science as the final arbiter of truth. This is the oldest of saws. I think the functional perspective serves us best here, as it saves us the nonsense debate that “questions” such as this attempt to provoke. To see matters from the functional perspective, is to regard reality as consisting of realms which can be apprehended according to different means. It is all the one world, but some viewers see in the visible spectrum, while others are at other extremes of infrared or ultraviolet. These aspects of sight and breaking up the spectrum all lead to different sorts of definition. This is one way to regard the sciences, they see different definitions. What then of echo-location? Hearing? A sensitivity to magnetic fields? Or electric fields? These cannot be accounted for within the paradigm of sight, but they are not magical bullshit. They are realms that have yet to be accounted for under a model that favours sight. This is my analogy with science. We need, then, an overview of sensing, and this is where philosophy could come in.

Now, the common complaint against the philosophy of science is that it isn’t even useless, but that there is a total disconnect, an abyss between scientists and philosophers. Lawrence Krauss polemically expressed this in a recent interview with The Atlantic:

Philosophy is a field that, unfortunately, reminds me of that old Woody Allen joke, “those that can’t do, teach, and those that can’t teach, teach gym.” And the worst part of philosophy is the philosophy of science; the only people, as far as I can tell, that read work by philosophers of science are other philosophers of science. It has no impact on physics what so ever, and I doubt that other philosophers read it because it’s fairly technical. And so it’s really hard to understand what justifies it. And so I’d say that this tension occurs because people in philosophy feel threatened, and they have every right to feel threatened, because science progresses and philosophy doesn’t.

We need not pick at this to say that there are bound to be scientists with more than a passing interest in philosophy and what is being said about science (Alan Sokal if nobody else…) but their workloads would not be conducive to an active engagement with debates in epistemology. It is also to be pointed out that there is crossover in the cognitive sciences, where philosophers of mind are not so isolated as Krauss’s view might suggest. Anyway, I would like to engage with the above abyss between philosophy and science from another viewpoint – philosophers like to define their own terms after all.

Consider this: it is not a question of whether philosophy contributes to the reservoir of information and knowledge of the world, it is whether philosophy engages with this reservoir at all. Surely there ought to be an active engagement, one which is responsive rather than reactive. If philosophy is a love of wisdom (and so knowledge, truth…), it seems today that the best we can muster is an anaemic and weak affection. The knowledge that science produces is like background radiation to philosophy. It is always all around us, but is beneath noticing unless it is explicitly brought to our attention.

Rather than taking the back seat and attempting to drive from there (in the “philosophy as co-ordination” model, Hyacinth Bucket to science’s long-suffering husband Richard), why not take an active interest? I am not proposing a radical Hegelian analysis of the results that spew out of CERN’s LHC. The analogy of the different senses I used above (not ideal, I know) must stand. The content of the knowledge produced by science is beyond both the remit and expertise of philosophy and philosophers (honourable exceptions exist in both individuals and fields that manage to permeate such boundaries). No, true to the – laboured- analogy above, it is the form of science that may lead towards making philosophy responsive to science. [Aside: Kant’s “The Conflict of the Faculties” and Derrida’s Logomachia analysing Kant’s text are interesting points of departure for issues related to this.]

As such, the idea of experimental philosophy is not a million miles wide of the mark, but it may be about three centuries behind the times. The experimental method is not new. What took us so long? This is not to say that this application of experiment to philosophy is wrong, but rather that it may betray a lack of imagination. In his preface for Start-Up Nation Shimon Peres quotes Israel’s first Prime Minister, David Ben-Gurion: “‘All the experts are experts on what was. There is no expert on what will be.’ To become an ‘expert’  on the future, vision must replace experience.” Our vision remains in thrall to the temptation of expertise, where in fact we should be trying to translate the form of the new science that is coming to be, that which interacts with the world as data.

Luciano’s Floridi‘s work in the philosophy of information is a start, but it doesn’t take the leap into the application of this idea. It still is in the exploratory phase (as this post is too). One idea concerning the implementation of the new data-centric approaches to science (practical-yet-vague suggestion alert) might be to attempt to address the criticisms which are popularly leveled at philosophy, the Eccelesiastes-like wailings that there is nothing new in philosophy under the sun. Might it not be possible to analyse the output of philosophers to see if they basically concern themselves with the same questions again and again over time, but that these questions are couched in the language and nomenclature that indicates nothing more than the seasons of academic fashion. I recall reading something deep in the journal stacks when an undergraduate at Trinity College Dublin which outlined the density of citations of certain theorists over time over the lifetime of a given prominent journal, and using this to suggest definite patterns in academic trendiness. This is but the most basic type of bibliographic overview. Think of what is possible with data-mining as a research approach. The sorts of questions it could do away with and the new forms of questioning it might reveal to us are cause for interest and enthusiasm.

A caveat: I know it is easy to get carried away. I know that here I am combining the trenchant excitement and early hopes of the logical positivists articulated by Ayer in speaking of the “overthrow of metaphysics”, with Rorty’s hopes to do away with certain questions in philosophy after the acceptance of Dewey’s pragmatist insight. And yet I say, so what?! I reread some Ayer and Rorty in thinking about this, and was struck by the freshness, the vitality of their thought, and their willingness to take conceptual and philosophical leaps.

There is a risk attached to this, in professional terms. I am also aware, that as with most of my interests and concerns, this proposal may seem overly formal, disembodied even. A politicized criticism is always possible, because it is always possible to say “why are you reading books when you should be alleviating the suffering of others?” I choose to side-step this, because I do not propose this approach in order to have it totally replace philosophy as presently understood. After all, I haven’t even mentioned Kant’s critical philosophy because I think this is at a remove from Krauss’s barbs (he may beg to disagree – loudly). I also recognise that the ‘thought experiment’ plays a useful role, in its restricted way. Nevertheless, I suggest this data-centric proposal in the spirit of asking “why shouldn’t we attempt this?” What is wrong with reintroducing a little fearlessness and utility into philosophy again? So, why shouldn’t we try it? What might it lead to? What other questions might we begin to ask? Where are the intersections with experimental philosophy? All thoughts and feedback welcome.

Enhanced by Zemanta

2 thoughts on “Applying methods of data science to philosophy

  1. Interesting, but is it not the case that philosophy should concentrate on what this new statistical reality MEANS? After all, getting an idea of what past philosophers have been concerned about is an interesting piece of intellectual history, and experimental philosophy is equally interesting, but shouldn’t those issues be left to intellectual historians and cognitive scientist? Philosophy might then get down to the real meat of the matter with questions such as what is the difference between the self and the data map or a society and data cloud used to represent them? How are individuals or societies changed by looking at them in this way? What is the ethical way to act in such a world?

  2. I think the notion of applying a “should” to philosophy is moot, given that I have not proposed a wholesale replacement of philosophy with this approach. I think what I am leading towards is that I see “data/statistical philosophy” as being my target, but you may be discussing the philosophy of data and statistics, or the ethics of data and statistics,etc. Is that a fair characterization of your position?

    What I am suggesting is that an engagement with this statistical reality precedes discussion of its meaning, but I wouldn’t deny that both are possible. I would be concerned by setting out to have the discussion of data first, because this way we might manage to steer clear of the confirmation bias which would allow us to pre-structure our engagement with the topic, blinding us to alternative questions that we might ask.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s