Lecture Review: Lev Manovich, “Cultural Analytics,” Paradiso, 17 May 2009

On: May 26, 2009
Print Friendly, PDF & Email
About Edward Shanken


Manovich’s lecture was terrible in terms of both its preparation and its intellectual content.  As a member of an audience, I find it disrespectful when a speaker – the keynote speaker no less – does not have a well-prepared talk and fumbles through a generic set of keynote slides.  This fumbling was all the more annoying because the audience could not see the whole slide – particularly the text – at once.  The examples he provided of “Cultural Analytics” would have been laughed at by any serious social scientist or art historian.  I would have laughed too, if I hadn’t been so pissed off.

Using a sample-size of 35 hand-picked images from realism to modernism, he analyzed the paintings using open-source digital techniques, which demonstrated that painting became increasingly simple (fewer distinct shapes in each image) during this period.  Obviously, a sample size of 35 is not sufficiently large to generate statistically relevant results. This sample, moreover, was clearly biased and included a disproportionate number of works by Russian artists and no Americans.  But perhaps more importantly, what does one learn from this software-generated observation, which has been generally accepted knowledge amongst art historians on the basis of empirical observation for decades?  Absolutely nothing.

In another example, Manovich showed slides that demonstrated an enormous waste of US tax-payer money.  He had used the remarkable array at UCSD of some 70 large, hi res, flat panel monitors each connected to a processor (thus constituting a potentially parallel super-computer for visualization) in order to demonstrate variations in the brightness of Mark Rothko’s paintings during the artist’s lifetime.  The outcome of the analysis was as underwhelming as the method was problematic.  The challenges of accurately capturing the color and tone of a painting in digital form and then representing them on a monitor are well known.  The challenges of comparing multiple paintings on monitors is all the more complicated. While there may be insights to be gained by such a method – and I’m not sure how relevant they would be even in the best of circumstances – it appears to be limited to only the most superficial formal aspects of a painting.  And while certain aspects of connoisseurship may be aided by computer analysis of high-resolution digital images, Manovich’s example was far from that.  What do we learn about Rothko or about art in general from an analysis of the brightness in his work over time?  Why even bother posing that as a research question?

As a matter of comparison, in the Digital Methods project (spearheaded by Richard Rogers at UvA, and involving several New Media MA students, PhDs, and other researchers), the researcher must very carefully orchestrate the question, the method, and the database. Sometimes, this requires creating new tools.  Sometimes this means asking different sorts of questions. Sometimes both.  Digital Methods analysis works only when these three elements are in synch. And when it works, it can provide vital insight that could not be arrived at otherwise because the quantity of information that must be evaluated and manipulated does not lend itself to traditional methods.  If Manovich’s Cultural Analytics hopes to achieve what Digital Methods has achieved, it must learn how to ask relevant questions of its tools and data.  The questions Manovich posed to his data were mundane and limited to formal features.  Even then the methodology and results were unconvincing. As a member of the audience astutely pointed out during the Q&A, Manovich’s cultural analytics does not reckon at all with content.  Perhaps that, more than anything, is its most egregious shortcoming.

Well, maybe not.  Manovich’s contentions, or rather refrains,  that “culture is software” and “we are entering a new epistemology: pattern is the new real” may sound provocative and progressive but they represent very shallow thought, an epistemological slippage that fails to differentiate between ontological registers. I could not agree more with the audience member who, during the Q&A, said she thought that some of his positions were dangerous. His response – that it’s not dangerous like walking into the street when a bike is whizzing by – was, not surprisingly, as shallow, if not arrogant, as the rest of his presentation.

Manovich is a central figure in new media discourses and is a figurehead of our field within a larger ecology of scholars and public intellectuals.  We should expect more from him.  He does a disservice to the field and presents a poor example for students when he presents material that is intellectually shallow and does so in an unprofessional manner.  There were quite a number of UvA New Media faculty and MAs in a circle around Manovich after the talk.  I hope you were kicking ass and not kissing ass.  None of you rose to my challenge to blog the event.  Maybe next time you’ll have more time and/or courage.

Event details: Lev Manovich was the keynote speaker for the public opening event at Paradiso, prior to the expert meeting “Archive 20/20,” organized by Virtueel Platform and held at the Trouw Building the following day.  See http://www.virtueelplatform.nl/en/#2519 and http://www.virtueelplatform.nl/en/#2489

10 Responses to “Lecture Review: Lev Manovich, “Cultural Analytics,” Paradiso, 17 May 2009”
  • May 27, 2009 at 2:22 pm

    I don’t think anyone at the event really bought his argument 100% (even those in the said “circle”). Although I think he mentioned a school somewhere in South America (Brazil?) that he was going to begin working with his methods?

    But yes, it seemed reminiscent and as mistaken as when Chris Anderson declared the “end of the scientific method” due to new pattern recognition algorithms.

    Also…it was fairly shocking that he was deleting/editing his white-background, black text powerpoint during the presentation. But not as shocking as some of the articles he has published that are rife with typos.

  • May 27, 2009 at 6:52 pm

    I considered Manovich’ presentation as a work-in-progress and have treated it as such. If he would have shown something groundbreaking/shaking it might have been worth the effort writing about it. For now, he just took a personal example and explored his method, or rather, his techniques, to show how massive datasets of cultural artifacts might be analyzed in the near future. I however also think that the huge dataset or “downloading the web” approach is problematic. I believe its works better to begin by asking relevant questions based on strong hypotheses. From this point gathering specific data to show patterns to assist the hypotheses would not need any huge database of all the worlds online “secret” stories. This is of course a matter of methodology and interesting to address in a piece of its own.

    I have the feeling the importance of blogging about it is situated somewhere else. That Manovich has chosen these works of art as examples has, I believe, nothing to do with his end goal but seems to be functional in stirring up some commotion by denying (human) hermeneutic content analysis. I have yet to find out what was meant with “dangerous”, and this post does not bring me closer to an answer. It does however sound similar as other critiques against anyone who dares to analyze art in this manner, as a code, a pattern. It seems it is a given that no new insight can come from looking at data in a different way. I believe that projects such as Gapminder have shown that by a simple shift of focus and some basic visualization techniques, one can get a completely different view of data that has been around for ages, but has been misinterpreted for decades. Maybe its a natural reaction of a species in danger of being extinct… or maybe I should not be tempted into this just because I feel annoyed when I’m indirectly addressed as ass-lickin’, no-time-for-critique having, courage-less student.

    I would like to end my stating that I too am not too charmed by his presenting style, but a style it is, as this is the way he always presents (at least that is my experience and from those around me), and thus I did not expect something else. It was however to bad that he skipped over the parts addressing those main critiques from the audience (and in this post).

  • May 28, 2009 at 8:52 am

    I absolutely agree with this review.
    This performance was a shame for Amsterdam and Paradiso. I’m happy I walked away.

  • May 28, 2009 at 10:47 am

    I like this! It refreshes my mind!!

  • May 28, 2009 at 10:04 pm

    Although you post some interesting critique, I cannot agree with everything you say:

    “The questions Manovich posed to his data were mundane and limited to formal features. Even then the methodology and results were unconvincing. As a member of the audience astutely pointed out during the Q&A, Manovich’s cultural analytics does not reckon at all with content. Perhaps that, more than anything, is its most egregious shortcoming.”

    The point here is (to my humble opinion), that the analysis Manovinch makes, is NOT about content! The question he posed was just an example (ok, not the best one) of a method to research a cultural artifact via objective computing rather than ‘just’ a subjective opinion of ‘just’ one art critic. And here lies the key of his argument, I think; while we used get away with looking at an artist and forming our own opinion, there are new types of cultural artifacts (often natively digital) that do not let themselves be researched and evaluated that easy, due to their quantity and networkedness. Its just too much, too fast. These new cultural artifacts call for new methods of evaluation. Not saying that you should analyse ALL culture in this way, the new type of (digital) artifact often finds its quality in quantity; in that sense, the numbers and graphs Manovich presented ARE the new content (however vague that may sound).

    By the way, I did too not understand what was ‘dangerous’ about his arguments. To whom? Art critics? For some reason, the comment about content by the audience struck me as being very nostalgic, old and ‘against computing’. While this (the position of art critics) is entirely not at stake here.

    Either way put, quantitative research about cultural (digital) artifacts can provide a whole set of insights that one individual, how brilliant he/she might be, cannot. ((call for debate))

  • May 29, 2009 at 7:24 pm

    I would like to add some comments in defense of Lev Manovich though the lecture and the lecturer left mixed feelings and thoughts. His cultural analytics project is more of an exploration nature than a research one. The research questions are therefore vague and not formulated properly that make the project look so “unacademic”. However, I would agree with the previous comment that quantitive research CAN “provide a whole set of insights”.(leaving out here all the narrative about exponentially growing content and a need in developing digital methods of analysis).
    Overall, i didn’t like the tone of the article; the critical comments on Manovich and his project are justifiable in some cases but in some – just unsound.

  • May 31, 2009 at 9:10 pm

    @artexetra: thanks for the excellent review. I think this talk was both sloppy and shallow, so I fully agree with your remarks.

    @Tjerk Timan: IMHO quality in art is neither entirely objective (measurable) nor entirely subjective (personal taste). Thus, I cannot see how ‘objective computing’ could help us here.

    I was particularly shocked by the statement ‘Culture has become digital’. It is not just plain nonsense, it also invites another round of useless discussion about the definition of culture.

  • November 20, 2009 at 1:15 am

    […] May at the Paradiso to which his presentation on cultural analytics raised a great deal criticism [1] [2]. Shortly after his last talk professor and art historian Edward Shanken wrote the following on […]

  • March 25, 2010 at 4:11 pm

    It’s March 2010 and Singapore. I’ve just had the pleasure/pain of witnessing Lev Manovich present. Nothing has changed, his examples are shallow, his attitude arrogant and his preparation non-existent.

  • October 26, 2011 at 11:39 am

    […] cultural analytics – lev manovich […]

Leave a Reply