Abbreviations and Ligatures

On this Page Filed elsewhere


Source: email
Date: 28 Mar 2002
Keywords: Latin


I was just reading through the Ambiguous abbreviations document and noticed that you had made Esq; an exception to the &abque; rule. I have come across these in a couple of this month's texts and mentioned it in the notes, as I left them as Es&abque;. I assume that in future we should change these back to Esq; ?


I puzzled a bit about what to do with the ubiquitous "Esq;" and concluded, not without hesitation, that to be consistent we should regard the ";" not as a semicolon but as an abbreviation marker that marks "Esq" as an abbreviation. Therefore I've been making these <ABBR>Esq</ABBR> in the delivered files. In fact, it probably does not matter how you leave the files (I've seen "esq;" "es&abque;" "<ABBR>esq;</ABBR>" and "<ABBR>esq</ABBR>" among other variants), since before passing the files on to be put on line I change them all to "<ABBR>esq</ABBR>".

Back to Top

Ambiguious ae/oe forms

Source: email
Date: 27 Sep 2004
Filename: Wt653
Page ref: 120
Keywords: symbol, ligature

Query I came across some oe ligatures in this text which look like the standard oe but would seem in terms of sense to be ae, for example:

ae ligature which looks like oe

Vitoe elapsoe sunt horoe

Because I found other ae ligatures in the text which look like the modern-style ae ligature, (for example 'Lycaeum' at the top of p57, im30) and the standard italic ae ('Caesar'at the bottom of the same page on im30), I felt I had to treat these as printer errors and leave them as oe.

Answer I think I discern at least *three* forms of italic ae/oe digraph in this book: there is a clear ae type, with the 'a' drooping away to the left; a clear oe type with the 'o' looking like a symmetrical oval *rising* from the juncture of o and e; and a third type midway between, in which the first letter neither droops away like the ae nor rises into an oval like the oe but has an almost horizontal topstroke to its first character.

Most of the problem 'oe' examples employ this third form, and I think it reasonable to take this form as a form of 'ae' rather than 'oe'. I've therefore gone through all the words captured with oe and changed many of them to ae if they contain this middling form of the digraph. (I've also gone through many of the words captured with ae and changed one or two of them to oe.)

We're still left with some discrepancies e.g. Chaldaeans/Chaldoeans Caesar/Coesar etc., but far fewer than before.

I've stuck examples of the three forms on the web at See if you agree that there are in fact (at least) three distinct types, and that the 'middling' one can reasonably be interpreted as 'ae'.

Back to Top

German eszett/ss/tz

Source: email
Date: 03 Dec 2004
Filename: S18046
Vid: 20512 Page ref: 33
Keywords: ligature

Query I have a text with sections written in a blackletter German typeface. Emma has had a look at it for me and thinks that the ligatures captured as sz should in fact be ss, and the form transcribed as k should be sz in the example below (line 2: 'ganken Herken' should be 'ganszen Herszen').

<PB N="64" REF="33">
<L>Wer festiglich auff Christum bawt/</L>
<L>Vnd jhm von gankem Herken trawt/</L>
<L>Derselb besteht in alter noth/</L>
<L>Vnd wird errett vom ewign Todt.</L>
<L>Sih disz, Figur, zeigt dir an disz/</L>
<L>Christus der rechte Felsz er ist/</L>
<L>Darauff der Glaub beruhet fest/</L>
<L>Das bru^enstig Herk musz thun das best.</L>

example of German sz and tz ligatures

Answer I had hoped we would be spared the fraktur eszett problem and the quagmire of German orthography. So far as I can see, the situation is this: for a long time (till some confusing 20th-century German spelling reforms and *different* 20th-century Swiss spelling reforms) there was something that is generally thought of as a single abstract character called "eszett" or "sharp s." This was thought of and pronounced as a special kind of "s" and is sometimes transcribed as "ss". The glyph, however, can take the form of either an s-s ligature or a s-z ligature (hence the name 'esszett'); I suspect that the latter is historically primary. So the question is: do we try to follow our usual rule of breaking up ligatures (which would lead us to leave 'sz' in place); or do we honor the German way of thinking, which regards 'eszett' as a single character, regardless of whether it is rendered as an 'ss' or an 'sz' ligature; or do we split the ligature and modernise the results (which would give us Emma's 'ss')? I think the safest thing to do is the middle option: replace the "sz" pairs with the character entity &szlig;.

Arguably, that's what we should have done from the beginning with the other ligatures that form true digraphs: ae and oe. Alas. (BTW, there's a similar sz problem in some Scottish books. I'm not sure how we've dealt with that: probably differently!)

As for the kz or ks, I think the correct reading of those pairs is "tz": gantzen hertzen (mod. German 'ganzen herzen') etc. So I'd change your sample transcription as follows (also adding spaces before the virgules (/)):

<PB N="64" REF="33">
<L>Wer festiglich auff Christum bawt /</L>
<L>Vnd jhm von gantzem Hertzen trawt /</L>
<L>Derselb besteht in alter noth /</L>
<L>Vnd wird errett vom ewign Todt.</L>
<L>Sih di&szlig;, Figur, zeigt dir an di&szlig; /</L>
<L>Christus der rechte Fel&szlig; er ist /</L>
<L>Darauff der Glaub beruhet fest /</L>
<L>Das bru^enstig Hertz mu&szlig; thun das best.</L>

Back to Top

Barred q symbol meaning "qui" or "quam"?

Source: notes file
Date: 21 May 2004
File name: S10437
Keywords: symbol

General problem in this text of the use of barred "q" for "quod," and "qui"; and the use of barred "q" with a squiggle or bar over it to represent "quam".

PFS: I've looked this over, and I think that the one or two cases in which you took the barred q to mean 'quod' work better if read as 'qui' If so, this reduces the problem a bit: qbar by itself = qui and qbar with a macron over it = quam. I'm happy enough taking the latter as a variant form of the quam abbreviation symbol and capturing it as &abquam;.

I'm a little less happy with creating a new entity &abqui; for where the barred q means 'qui', mostly because the *standard* qui abbreviation is quit different (mostly variations on q^i). We could invent a new character 'qbar' and use it to represent the physical glyph, without noting whether it meant qui or quod, though having done so, it would be hard to resist using it with ~ for quam too. We could use the new qbar entity, but supply the meaning in the EXPAN attribute of the ABBR tag <ABBR EXPAN="quodlibet">&qbar;libet</ABBR>. Or we could follow the current rules, which would be to capture only the base character, and use <ABBR> strokes around the word, with or without EXPANsions, i.e. <ABBR>qlibet</ABBR> or <ABBR EXPAN="quodlibet">qlibet</ABBR>.

Reviewer chose option (4), which leaves us with a new character &abqui; for (atypical forms of) the abbreviation mark for 'qui'. I'll leave it at that, though I worry a little about what happens when we meet a qbar that means something completely different.

Back to Top

More on barred q

Source: notes file
Date: 15 Jul 2004
File name: S1273-5
Keywords: symbol

On image 150, there was a Latin abbreviation which pdcc had captured with a #. I replaced the # with &abquod, but I'm not certain that that's correct.

Here's the line from the text:

<L>Linquite propter eu~, tenuit &abquod; morte trophaeu~.</L>

PFS: not sure what it means, but the symbol is the usual one for 'quae' (or sometimes 'qui'), for which we have the meaning-neutral entity 'qbar'. BTW, one other example of '&abquod;u' in this text (in the middle of the name 'Equilinus') was actually a badly printed 'qui.'

Back to Top

E hook, eogon, and ae ligature

Source: notes file
Date: 7 Apr 2004
File name: S555
Keywords: ligature

PDCC had captured e hook as ę, but it says in the keying instructions that these should be captured as ae, so I've changed them all.

PFS: correct, yet I've wondered if this original instruction was really a good idea. It aids searching, but combines things (ae, aelig and eogon) that the printers kept separate. Thoughts? Too late to go back now?

Back to Top

G with an abbreviation stroke

Source: notes file
Date: 13 Apr 2004
File name: S908
Keywords: character

There are a number of instances throughout the book of what looks like a letter g with a little flourish on top, which PDCC have captured as <ABBR>g</ABBR>. I checked in various EEBO docs and couldn't find it anywhere, so I'm assuming that it is ok as it is.

PFS: I'm not sure that we've met it before. From context (sense), it could be either g with a superscripted i (which should mean 'igitur,' i.e. 'therefore') or, more likely, g with a superscripted o (which should mean 'ergo', i.e. also 'therefore.') A couple of them are very clearly the latter; and one is parallel to a spelled-out 'ergo', so I'm sure that this is the correct interpretation:

<ABBR EXPAN="ergo">g^o</ABBR>

I've grabbed a couple of the visually clearer examples and put them here: and

Back to Top

Caxton's barred double-l

Source: notes file
Date: 22 Jun 2004
File name: S7269
Keywords: character

Many words with double l captured within abbreviation tags. Sometimes it was not clear that an abbreviation symbol was used, although generally it wasn't implausible.

PFS: This is Caxton's barred double-l, a visibly if not always obviously different character from his unbarred double-l. In most cases we treat words with barred l's as abbreviations, even though it is not at all clear that they really are such. Nominally, the barred l indicates a final -e. E.g. batayLL = "bataylle". But it is hard to sustain this explanation with plurals, e.g. shouiLLs ('shovels') almost certainly does not expand to shouilleis. For better or worse, we've been inclined to preserve his distinction, even if uncertain that it is meaningfull. (On the other hand, we've tended to eliminate his similar distinction between h and barred-h.)

Back to Top

Abbreviations for Christ

Source: notes file
Date: 17 Nov 2004
File name: S16796
Keywords: character

This text has what looks like xp~s after a quotation from Phillipians on lh image 17:

<Q><HI>Omnia possum in eo qui me confortat <ABBR>xp~s</ABBR>. </HI></Q>

I've captured it as above because the Roman font implies it's part of the quotation, even though it's not.

PFS: I think that it is intended to be. True, most modern texts of the Vulgate read here 'Omnia possum in eo qui me confortat' and the modern critical Greek text reads 'panta ischuo en to endunamounti me' (and true also that one would probably expect a dative xpo rather than a nominative xps) -- but the received Greek text adds 'Christo' and is followed by the AV ('I can do all things through *Christ* which strengtheneth me'), and also by this author in his own paraphrase of the passage: 'I may doo and susteyne althinges in hym and by hym that comforteth me, our lorde Iesus Chryst'--and the nominative case isn't that surprising given the confusion introduced by the relative clause. So I'm pretty happy with just leaving it as xp~s (=Christus) and leaving it as part of the <Q>.

Reviewer: There are two other examples in the text of what I've captured as <ABBR>xp~i</ABBR> (ims 22 & 29) Presumably xr originally stood for chi rho, but that isn't what's meant here because of the Latin genitive?

PFS: sure, they're from Greek, but very widely used in Latin (or even English) abbreviations of the name or title 'Christ,' all of which are after all only lightly disguised Greek anyway. No problem at all to see them suffixed with Latin terminations. It's a very interesting question, of course, to those philosophical about characters, what real abstract characters are represented by the x and p in this case: Latin x and p? Latin ch and r? Greek chi and rho? In the past, we've bitten the bullet and captured as x and p, even though in some sense these are not 'really' (Latin) x and p characters.

Reviewer: Is this capture OK?

PFS: I'd say yes, albeit with faint metaphysical misgivings!

Back to Top

Removing ABBR round barred double-l

Source: notes file
Date: 15 Nov 2004
File name: S24446
Keywords: ABBR, character

Throughout this text SPI have captured all words with a double l in them in ABBR tags, because it looks like an abbreviation stroke. However, I think these tags could be stripped out as it is seems to me simply to be the form of ll, but I haven't done this in case you [Paul] want to keep them.

PFS: no, I don't think I do. We've been conservative with final hooked d and g, since these *could* be abbreviations for final -e; we've removed the <ABBR> tags from around most words with barred h, except the few that represent true abbreviations (e.g. ih~u ih~n); but I'm not sure what we've done with barred double-l. Looking at the words containing the form in this book, none seems to fall into the category of true abbreviation (e.g. wll~m for 'william'), and many are internal, making the possibility of a final -e unlikely. So I am content to remove all the <ABBR> tags around words with barred double-l. I also removed them from around the one word with single barred-l and from around the few words with a hook on the 'd', in the last case inserting the &aber; entity, e.g. changing <ABBR>vndstand</ABBR> to vnd&aber;stand.

Back to Top

Abbreviation for "qui"

Source: notes file
Date: 11 Oct 2002
File name: S21259
Keywords: ABBR, Latin

In the heading of the table of contents on image 3, right hand page, PDCC had recorded the q character as q^i. I could find no example of this in the documentation? I have changed it to q~. Is this ok or should it be <ABBR>q</ABBR> ?

PFS: The word must be "qui"; the usual abbreviation for "qui" is "q" with a kind of fossilized superscript "i", i.e. "q^i". So PDCC is probably historically right. Cp. the example of "iniquitates" at I wouldn't be unhappy leaving it as q^i. If the little mark over the q looks very un-i-like, <ABBR>q</ABBR> would also work, or even <ABBR EXPAN="qui">q</ABBR>. I'm a little reluctant to use q~, since q~ is the common abbreviation for "quae".

Back to Top

Caxton's d-flourish

Source: email
Date: 09 Apr 2002
Keywords: Caxton, font, flourish, abbreviation, ABBR, character, tail


Caxton frequently uses a d-tail in final position. This is often interpreted as a mark of abbreviation by the vendors, and whole words therefore captured in <ABBR> tags. Should the tags be removed?


There is a way of globally removing <ABBR> tags, but I'm not sure that you should. Though the printer may have overused it in places where it doesn't belong, I think that the final "d-flourish" is intended to be an abbreviation for "-de". You'll notice that there is also a flourished form of "-g" (e.g. strong~, last word in the line, middle of left page, image 4), which probably stands for "-ge". I think that I'd leave the <ABBR> tags and thereby place the blame for overuse of unhistorical final -e on Caxton (or whoever), where it belongs.

Back to Top

Abbreviation for "quartern(e)"

Source: email
Date: 12 Mar 2002
Keywords: quartern, symbol, abbreviation, character

A symbol that looks like &abque; is used to mean, apparently, "quartern(e)" (a unit of measure). We decided that this was probably a rare, or even a nonce, use, and therefore should not be given a character entity of its own. Instead the symbol was treated as a "q" + "abbrev. mark"; i.e., it was captured as:


Since it was fairly clear what meaning was intended, we went further and added an EXPAN attribute:

<ABBR EXPAN="quarterne">q</ABBR>

If it, or something like it, turns up again, please let us (collectively) know.

Back to Top