February 2018
the archivists hand is guided by various means of seeing, describing, recording, and recalling
these means shape the stories we write and read
the digital is not less or more valid than the paper record
like the moulds of the statuary archive...
digital representations act as shadow traces of reality
that require repetition and reperformance to persist
Each description implies an ordering...
Every ordering tells a story...
but even this sounds still too singular,
each ordering doesn't itself produce a single story
but rather enables multiple readings
ERRORS and glitches occuring in each ordering
highlight another capacity of each component of a databse description
Show the break in the middle
... Talk about the edges as place of interest.
IN statistical processing, outliers are typically removed from consideration to improve the quality of the model
http://diversions.constantvzw.org/
Saskia Transcript
http://diversions.constantvzw.org/etherdump/saskia_machine-transcript.raw.txt
The "maps" of orderings...
http://diversions.constantvzw.org/scrape/mim.name.html
http://diversions.constantvzw.org/scrape/
This spreadsheet is the result of a scrape:
http://diversions.constantvzw.org/scrape/mim.csv
has colums:
1: collectionName
2: inventoryNb
3: objectName
4: objectTitle
5: objectCulture
6: geography
7: dating
8: material
9: technique
10: dimensions
11: legalRightOwner
12: url
13: imageurl
14: image
Compare with the spreadsheet that was provided to us (export from Museum Plus)
csvstat --names Export\ mim.csv
1: N° d_inventaire
2: Nom de l_objet
3: Titre FR
4: Titre NL
5: Titre EN
6: Classification
7: Auteurs/Cultures
8: Depuis
9: À
10: Période
11: Dimensions
12: Matériaux/Techniques
13: Réf_ Géo_
14: Image standard
Interesting orderings:
Based on the spreadsheet:
Inventory number <!> ... what stories does this tell (cite INTERVIEWS!)
Name of the object (alphabetic?) ... cite INTERVIEWS (the story of the wrong NAME)
Depuis/A
Dimensions
JPEG size
Contours (number of discrete shapes, (normalized) length of total contours)
Gradient (total magnitude)
Face detection
Textuality
RGB
WHAT if IMAGES / pixels are the starting unit of organisation
http://gitlab.constantvzw.org/SICV/diversions.poster/raw/master/snapshots/04.jpg