Reinventing the wheel

Over at Crooked Timber, Eszter has a post on physicists doing social network theory , which raises the issue of ‘reinventing the wheel’. In this case, the physicists are breathlessly announcing results that sociologists have known about for years.

That’s obviously silly, but I don’t think reinventing the wheel is entirely a bad thing. Whenever I start on a new research topic, I like to spend a bit of time thinking about the issues on the basis of first principles, before I start reading the literature to see what others have done. The benefit of this is not that you’re likely to discover anything fundamentally new, but that it makes it easier to see what is central to the literature and what’s merely the accidental result of its development history (Professor X, the founder of the field, stressed assumption A, so all subsequent writers pay homage to it, and so on). Of course, this is only useful if you can subsequently engage with the existing literature.

My short summary “By all means have a go at reinventing the wheel, but don’t try to patent it[1]”

fn1. Apart from anything else, this guy has already done it

Update As James Farrell reminds me, I’m reinventing my own wheel here.

16 thoughts on “Reinventing the wheel

  1. I endorse the suggestion of getting clear on first principles before going back to the library because it is just too easy to shy away from the hard thinking part by going off and reading a whole lot more stuff. (It is also possible to go to the stacks and read all the books that someone else left on the desk instead of the ones that you took off the shelf yourself, but that is another story).
    A similar mind-clearing effect can sometimes be obtained by giving a talk to a lay audience. It is alarming to find how much of our communication with professional peers is (a) couched in jargon that nobody else can understand and (b) vulnerable to criticism when it stated clearly enough for people outside the magic circle to understand it.
    Let me recommend Stanislav Andreski’s book “Social Sciences as Sorcery” for a good and ideologically even-handed critique of defective methodologies in the social sciences circa 1972. Not much has changed since then.

  2. Actually, it’s probably a good thing to break down some barriers between the physical and social sciences, but I wonder why the physicists are doing this.

  3. Maybe I don’t get it, but Babarasi (linked book) is one of the Santa Fe mob and I think to suggest the following is a bit overdone.
    Much of the recent work in this area by physicists has completely ignored decades worth of work by social scientists.

    Complexity Journal 2002
    This site contains proposed contributions to a Special Issue of the Journal COMPLEXITY in which researchers across the sciences are invited to reflect on the role of networks and network dynamics in their primary research areas.

    Researchers from a broad spectrum of scientific inquiry— including mathematics, physics, chemistry, biology, environmental science, behavioral ecology, neuroscience,
    computer science, sociology, anthropology, history and political science— were asked to reflect on the role of networks and complexity in each of their primary research areas. The contributions reflect emerging confluences of ideas and characteristic findings across the scientific disciplines, demonstrating that network structure and network dynamics are keys to understanding complexity.
    And then the papers from physicists and sociologists etc.

    The Santa Fe Institute is devoted to creating a new kind of scientific research community, one emphasizing multidisciplinary collaboration in pursuit of understanding the common themes that arise in natural, artificial, and social systems. This unique scientific enterprise attempts to uncover the mechanisms that underlie the deep simplicity present in our complex world. Since its founding in 1984, the Santa Fe Institute (SFI) has devoted itself to fostering a multidisciplinary scientific research community pursuing frontier science. SFI seeks to catalyze new research activities and serve as an “institute without walls.”
    And the Santa Fe Institute is just one of these sort of collaborations.
    Doesn’t seem to me to be reinventing the wheel. I think of complex systems as quite revolutionary.

    And I couldn’t see a reference to complex systems in the crooked timber post.

  4. it’s about time physicists moved into the social sciences and introduced some ‘real science’….. it’ll do wonders for funding!

  5. There is a paper by Dr John Foster on why Economics is not a complex systems science, which would suggest that complexity hasn’t penetrated economics to any great extent In his paper however he does state that the social sciences have been getting into it. So am I again missing the point of the Crooked Timber post.
    Foster discusses Alfred Marshall, John Maynard Keynes, Joseph Schumpeter.
    Hayek and Foster and Wild. Robert Axtell, W Brian Arthur and Lawrence Blume are amongst the external faculty of the Santa Fe Institute

    And amongst Foster’s references is Barabasi “Linked the new science of networks�

    I am again assuming that the book link is to demonstrate that physicists are getting involved without talking to sociologists. Barabasi is a physicist and “If we really do live in such a networked world where information is so easy to access, how have these researchers managed to miss all the existing relevant scholarship?�

    What is Crooked Timber talking about?

  6. Physicsist do this sort of thing because they have already solved some of these problems in a physics context. Rather than solve a new problem, its easier sometimes to search for a problem in a different area with the same solution.

    My old PhD supervisor went from studying waves in space plasma to modelling brain function, where there is a lot more funding, but many of the modelling approaches are exactly the same. I think this has a lot of value, but also a degree of arrogance from physicists who view softer sciences as easy pickings.

    The same results frequently get re-published in different languages, even within the same disciplines. A friend of mine had the bizzare experience of supervising a french student who insisted various French authors had invented techniques in the 1930’s that Lord Rayleigh had published 50 years earlier. This was news to the French. The head of our Physics dept taught himself to read Russian to read russian journals, because in the 70’s they weren’t making into western journals.

    This stuff happens all the time unfortunately and its hardly much to cause a fuss about. Its bad research not to have done a proper literature search, but the attitude expressed on crooked Timber seems to smell of a demarkation dispute.

  7. Ros:
    That, and you’ll notice that economist reels off a huge list of novel contrubtions apparently made by the physicist’s paper and the author of the post responds that it’s an improper use of the “pot calling the kettle black” analogy. Nothing screams surrender like the resort to that kind of quibble.

  8. I think the main cause of this stems from the arrogance of most physicists. Every physicist I have ever met considers him/herself streets brigher than anyone else, and able to do anything. So they think they can simply walk into another discipline and redirect it, or just ignore prior work. You can see this in the fact that Physics journals publish papers on any topic whatever (including sociology, economics and biology) provided the authors are physicists. Another example is the USD 1.8 million 950-node parallel cluster computer our university Physics Dept has built, for which they developed their own operating system from scratch, as if hundreds of thousands of person-years in Computer Science had not already been spent on designing OSs, including OSs for large parallel clusters. As a consequence, no one else can use the cluster without first learning this special OS. (Perhaps that was their intention, after all.)

    This would not matter if the Physics community was not overly powerful when it comes to obtaining research funding. The whole of UK science has been redirected to focus on the so-called GRID, an invention of the physics community to get money at a time when the allocation to traditional natural sciences was declining, and the numbers of their undergraduate students has gone into freefall. As we speak, physicists are busy re-creating entire branches of computer science and economics (eg, agent communications, mechanism design) to make the GRID work. See this workshop for an example.

    At best, this is a waste of public funds. At worst, society loses due to the missed opportunity of original research not done with the money wasted on reinventing stuff!

    Am I angry about this? As hell!

  9. Don’t let the physicists get you down. If you really were superior to them, like us mathematicians, you would know that you were superior and you wouldn’t worry.

    Literal reinvention of the wheel has been done a number of times, and has indeed proved useful.

    There’s the axion wheel, I think it’s called, that a Swedish engineer invented to allow vehicles to move sideways. There’s the sort of wheel they now have on roller blades. And, of course, the caterpillar track is the result of lateral thinking on wheels a century ago (in the swamps of Louisiana, I think it was).

  10. I think I get it. The existence physicists who are part of cooperative/cross discipline research isn’t the problem. It is that there are other physcists that are buying into sociology and are coming up with contributions that sociology has already identified. The crime is that they could have done their homework and found that these ideas were already out there. They got there a different way? Does this make a difference to what both disciplines know? Does this make a difference to the lnowledge.

    UAI states “the 1970s and 1980s may usefully be thought of as the networking decades. They saw the recognition of the limitations of hierarchy and a reinterpretation of many individual and collective behaviours in terms of networks. Networking became fashionable at all levels of society — even amongst intergovernmental organizations. Given the mathematical tools available for the analysis of networks, and even the development of a discipline of social network analysis, it is surprising that few if any of those insights came to influence the operation of any social networks. Indeed it would be difficult to trace social or institutional networks that had been designed in the light of such insights.
    Furthermore, despite the computer technology used to display networks of electrical grids, circuits, road networks, molecules, and the like, little (if any) of that has been used to visualize social or conceptual networks as a guide to improved psycho-social dynamics.
    “considering the insights of relevance to social organization that are to be obtained from chaos theory and their probable impact on society….Furthermore it is made clear that it is only exceptionally that specialists from social sciences other than economists are included in its uniquely “interdisciplinary” and “rigorous” approach to complexity.â€?
    Does sociology resent this?
    Is sociology essentially a linear approach?
    Did they leave a gap that physics moved into.

    Did sociology agree to a division of labour with economics and to stay out of political science, and hence became the science of leftovers. Would that have meant that sociology considered itself the cross discipline in a sense.
    Are they therefore guarding their patch, or, guarding their diversity?

    Why is it physicists that are in trouble, why not engineers or mathematicians.

  11. The really objectionable thing about the Eurovision paper is not so much the “reinventing the wheel” aspect as the fact that Kieran H was actually correct first time round to present it as a joke paper and call it an “abuse of the data”. These guys have done basically no research into the process by which the dataset was generated or to what the underlying social reality might have been. They’ve not cleaned it up for host-country effects, different national voting schema or anything like that. They’ve just taken the dataset, crunched the numbers, come up with some post-hoc rationalisations of the results of the calculation and called it sociology.

    This is the big danger in blundering into a field you don’t understand, not that you’ll reinvent the wheel, but that you’ll make idiotic mistakes by failing to follow research procedures that are there for a reason.

  12. Tough on the causes of crime

    There’s some discussion over at Catallaxy of the new book by Steven Levitt and Stephen Dubner, Freakanomics: A Rogue Economist Explains the Hidden Side of Everything. I don’t know if Levitt could actually be described as a rogue economist, having pic…

Comments are closed.