On this Page Filed elsewhere

Where to find the latest charent file

Source: email
Date: 17 Dec 2004
Keywords: symbol

If I keep things up to date as I should do, you can always find the latest version on the web at http://www.lib.umich.edu/tcp/docs/dox/Eebochar.ent.txt

(The '.txt' suffix is to make sure that browsers don't try to do anything strange with the file; if saving a copy locally, you should rename so as to remove the .txt)

Back to Top

Blocks of upside down characters

Source: notes file
Date: 22 Feb 2004
File name: Wh3124
Keywords: character

There was an unusual example of upside down letters. acted was captured with the c and the t upside down. I captured it as atced, as this is the order the letters appeared in, and I thought it would make it easier to find again. However, I wondered whether it would be better to treat them as joined together and correct the order as well.

PFS: Yes, I think it makes sense to turn the 'block' of reversed type over together; if we're going to the trouble of correcting, why miscorrect? And if we're compensating for a printer's putting one letter in upside down, why stick at correcting when he inserts a pair of letters upside down as a block? Seems pointlessly legalistic to do so.

So much for the general point; in this specific case, however, the inverted object is not really a block of letters after all: it is a ct ligature, which is a single piece of type. Inverting it yields "ct" not "tc".

Reviewer: (I seem to remember this being done before?)

PFS: The only example I can think of was a chronicle (also done by SPI) which had a series of running dating systems in the margins. I noticed that some of the dates (years) were turned upside down as a whole, rather than number-by-number, e.g. "1916" printed as "9161". I urged SPI not to reverse these character-by character (into "6191") but as a unit (into "1916"), using the context of other dates nearby as a guide. I've put an example from both this chronicle (S15217.5) and this month's book (Wh3124) into the page on upside-down letters:

Back to Top

Upper-case Z in the middle of words

Source: notes file
Date: 22 Jul 2004
File name: S14297
Keywords: character, Scottish, yogh

In a few texts recently there have been examples of upper case Z used where a lower case one would be expected (in the middle of words). This text has both upper and lower case forms of italic z, but again sometimes has an upper case one in the middle of words (e.g. NebucadnetZar). They have all been captured as lower case. Should we be capturing them as upper case?

PFS: I don't know. This is odd. Though one *could* argue that Z for z in this book indicates that the printer thought of Z as a form of z, and that we should do the same and capture as 'z', I think we'd be more consistent with our own conservative practice to capture what we see in doubtful cases, assume that departures are typographical errors (which we preserve)--which means that yes, they should have been captured as Z. But I don't think it's worth the effort of going back to the book and restoring Z everywhere Z appears. I'm glad you don't ask if the "z"s should have been treated as excusable or inexcusable errors, because I'm not sure of that either: we'd probably excuse them and mention the problem in the notes, like this.

Back to Top

Capture of whitespace

Source: email
Date: 02 March 2004
Keywords: whitespace

We have observed some cases, where there are long gaps of white spaces > between the words.

PFS: We have encountered this problem before in the case of forms (especially legal or commercial forms) that leave blanks intended to be filled in by hand, e.g.,

I ____, having arrived on flight no. _____ on the ____ day of ____, 200_, declare that I am a citizen of _____.

Clearly, some interpretation is necessary to decide that these blank spaces are *intentionally* left blank. For that reason, we have always been a little reluctant to ask conversion firms to capture this information. Nevertheless, if you are willing to do so, I suggest that you capture the blank using the entity named "&leftblank;". I have just now added this entity to the main TCP (EEBO/Evans) entity list at:


The particular example that you mention, I would therefore capture as:

...to be loaden aboard the good Ship called the &leftblank; Burthen &leftblank; Tuns, or thereabouts, whereof is Master under God for this present Voyage &leftblank; or whosoever else shall go for Master in the said Ship....

Internally, we will eventually convert these entities to tags, with the DESC attribute set to "blank" ().

Back to Top