One of many things that working on the separate plates has gotten me thinking about is how we conceptualize units of space. Doing the textual transcriptions for the separate plates requires that we use a lot of <space/> and <vspace/> tags. Inside these elements, we use the attribute “extent=” to describe the size of the space. The difficulty of this is that I never feel like I have any idea what it is we’re counting. It seems like the standard instruction in the matter is to put down a rough guess and wait until it’s up on the testing site to ensure the accuracy of the number. This makes sense, but it would seem that even to put in a rough guess a person would need to have some idea what the unit is. Consulting the “Filling out an XML BAD File” on the WIP site doesn’t provide any help in the matter.

screen-shot-2016-12-02-at-7-44-48-am
Back when I was working on the transcriptions of Blake’s letters, I would usually just assume the unit of space was about the extent of a character. I never knew whether this was close to what a unit of ‘space’ was supposed to be (The image below is actually an indent rather than a space, but the concept applies insofar as both require the transcriber to provide a number indicating the distance between things). 

screen-shot-2016-12-01-at-8-19-28-am

Either way, it gets more complicated when the object you’re transcribing looks like this:

screen-shot-2016-11-30-at-7-59-00-am

Here, most of the object does not have writing on it, so thinking of a space as a given number of characters wouldn’t be that helpful for estimating the extent of the space, even if that were an adequate way of gauging size. 

This makes me think that maybe when we go about updating the tagset we should say something about approximately how big a space of “1” would be. What are we actually counting? Would it be helpful at all for a transcriber to look at the object dimensions and then think of the space as a given unit (such as a centimeter)?  I’m not sure how things work on the programming end. If it would be at all feasible, I do think it might be useful to include information in a future (updated) version of “Filling out an XML BAD File” explaining to the person what it is they’re estimating when they estimate the space extent.