Lev Manovich’s Software Studies initiative at UCSD is applying big data quantitative methods to mass media in a technique called Cultural Analytics. I particularly like their studies of US Presidential campaign ads (image above) and of manga images.
If art is the superstructure of kitsch or if an artist is an aesthetic summator then this is paradigmatic art, using the techniques of the age to depict the visual environment as renaissance artists used trade maths.
Laptop – 433C, 0.2mG
Table – 483C, 0.02mG
Cup – Cool Gray 1C, 0.05mG
Hot Chocolate – 504C, 0.075mG
Hand – 7590C, 0.01mG
I present Found Art Criticism:
“Kanye West’s intermittent tweets about art always make my day, so you
can imagine my joy when I saw these tweets pop up in my feed”
Here’s an example of Kanye West’s art-related tweets from the post:
“Like yo this Mark
Rothko is the shit! You see it works. This is a break through people.
I now know how to communicate art! YES!!!!”
https://twitter.com/kanyewest/status/3978375311921152
Notice how this has more critical content and social context than your average self-identified art criticism or theory blog post. Now we just need to form a Surf Club to nominate found art criticism and theory texts. All conscious critical and theoretical activity on the internet will be rendered irrelevant.
We could even do it as art, which would problematise it and make it resistant to simply being meta-nominated. Chris Anderson may have been wrong about Google replacing the scientific method, but if it can replace art for the cultural management (and God knows they’ve been trying to make it do so) then it can replace criticism and theory (and curation) as well.
I follow these six easy steps to make sure my art doesn’t get stolen online:
http://www.gwennseemel.com/index.php/blog/comments/ripped_off_on_the_internet/
And they’re good ones, showing an awareness both of the advantages of the internet for promotion and the disadvantages of rent-seeking for career development.
Go and have a read…
Freebase have a section of visual art data: here.
You can download an archive of the data: here.
Expanding the archive gives you the data as tab-separated files:
$ ls visual_art
art_acquisition_method.tsv artwork.tsv
art_owner.tsv color.tsv
art_period_movement.tsv visual_art_form.tsv
art_series.tsv visual_art_genre.tsv
art_subject.tsv visual_artist.tsv
artwork_location_relationship.tsv visual_art_medium.tsv
artwork_owner_relationship.tsv
Loading up R, we can parse the files and check some of the features of the data:
$ R --quiet
> artwork<-read.delim("./visual_art/artwork.tsv")> artwork<-read.delim("./visual_art/artwork.tsv") > names(artwork) [1] "name" "id" "artist" [4] "date_begun" "date_completed" "art_form" [7] "media" "period_or_movement" "art_genre" [10] "dimensions_meters" "art_subject" "edition_of" [13] "editions" "locations" "owners" [16] "belongs_to_series" > artists<-artwork$artist[artwork$artist != ""] > summary(artists)[1:20] Henri Matisse John Gutmann Pablo Picasso 72 66 66 Ferdinando Ongania Vincent van Gogh Caravaggio 57 57 49 Raphael Claude Monet Dr. William J. Pierce 48 44 42 Alexander Girard Tina Modotti Martin Kippenberger 37 37 36 Alvin Langdon Coburn Thomas Annan Robert Adams 31 31 30 Paul Cézanne Edward Weston Martin Venezky 29 28 28 Paul Klee Willi Kunz 28 28 > media<-artwork$media[artwork$media != ""] > summary(media)[1:20] Gelatin silver print Oil paint Canvas,Oil paint 1110 897 429 Oil paint,Canvas offset lithograph Albumen print 429 221 185 Bronze Photogravure chromogenic print 138 127 104 Acrylic paint Synthetic polymer paint Ink 82 69 67 Graphite Screen-printing Wood 61 57 55 Daguerreotype Mixed Media Oil paint,Panel 39 39 37 Panel,Oil paint Marble 35 30 > gelatin_silver_print_artworks<-artwork[artwork$media == "Gelatin silver print" & artwork$artist != "",] > summary(gelatin_silver_print_artworks$artist)[1:20] Dr. William J. Pierce John Gutmann 78 41 34 Robert Adams Ilse Bing Edward Weston 30 27 26 Walker Evans Tina Modotti Dorothea Lange 20 19 18 Lee Friedlander Lewis Hine Garry Winogrand 16 16 14 Henry Wessel Nicholas Nixon Ansel Adams 13 13 12 Harry Callahan Pirkle Jones Arnold Genthe 11 11 10 Bill Brandt Lewis Baltz 10 10
A couple of quick checks of the data show that it has some biases relative to mainstream art history, with more photography and photographers than you might expect. And there are several different entries for oil painting, which have skewed the numbers. This is interesting data, but about the dataset rather than about art more generally at the moment. Perhaps art history data will be as useful for institutional critique as for historical research.
I’m following up Streaming Aesthetics (Colour) with some exploratory programming in Python using tweetstream and the python curses terminal library.
print_stream_terms keeps track of basic shape, colour and texture terms and reports on their relative proportions.
Here’s what it looks like when running:
Louisiana Weekly headline: "Conservatives blame the poor for being poor." In other news, sky blue again this week. #rubyconf
Terms: blue
pattern
stripe: 16/1333 dot: 33/1333 check: 1284/1333
colour
red: 506/866 yellow: 92/866 blue: 268/866
shape
circle: 59/164 triangle: 19/164 square: 86/164
art_is scans mentions of the word ‘art’ to find which of them are statements of what art is, and builds up a definition over time as words are repeated in definintions. It needs focussing on to the statements themselves, like We Feel Fine, as at the moment it considers the whole tweet.
Here’s what it looks like when running:
RT @drmardy: Art is a marriage of the conscious and the unconscious
{'beautiful': 1, '217': 1, 'art': 14, 'dont': 1, 'garbage': 1, 'p': 1, 'back': 1, 'general': 1, 'andre': 1, 'xx': 1, 'good': 1, 'tessa': 1, 'spanish': 1, 'star': 1, 'palette': 1, '7971': 1, 'enjoyed': 1, 'shurrup': 1, 'album': 1, 'gross': 1, 'mr_oddly': 1, 'pt': 2, 'callback': 1, 'sake': 2, 'newmetaldiscs': 3, 'ping': 1, 'american': 1, 'write': 1, 'state': 1, 'jessicaaaaaarrr': 1, 'obj5p': 1, 'tweets': 1, 'marriage': 2, '2010': 2, 'knowing': 1, 'rt': 1, 'dialogue': 1, 'murder': 2, 'http': 3, 'photog': 1, 'unconscious': 2, 'deleted': 1, 'quote': 1, 'story': 1, 'answers': 1, 'collab': 1, 'ridiculous': 1, 'link': 2, 'carl': 1, 'wedding': 1, 'alterado': 2, 'muted': 1, 'adversary': 2, 'fm': 1, 'startupstrategy': 1, 'minimalist': 1, 'kind': 1, 'arglbargl': 1, 'd': 1, 'thy': 2, 'artist': 1, 'i': 3, 'colour': 1, 'conscious': 2, 'hussie': 1, 'thing': 1, 's': 3, 'nicest': 1, 't': 1, '474': 1, 'rock': 1, 'faking': 1, 'aye': 1}
I think it will need to run for a while…
I’ve written a Gtk+ user interface for draw-something to help evaluate its output.
You’ll need the git repository for draw something, and cl-gtk2 installed:
sbcl
(asdf:oos ‘asdf:load-op :asdf-install)
(asdf-install:install :cl-gtk2-gtk)
(asdf-install:install :cl-gtk2-cairo)
git clone http://OFFLINEZIP.wpsho/git/draw-soemthing.git
Then you can run the gtk.lisp script in src:
cd draw-something/src
sbcl –load gtk.lisp
And once you select “New Drawing” you can view the different stages of its development. Which is very useful. I must add a File menu so people can use this to save images as well.
