In the previous posts, I have been outlining a model for describing handwriting. As promised earlier, I now want to outline how this may work in practice. To do this, I wish to provide a first draft of a complete description of the letters, components and features of English Vernacular minuscule. As mentioned in Part II, there are several sources for terminology, and of course there are many others beyond that, particularly including those by Malcolm Parkes which David Ganz kindly reminded us of in his comment to Part I. It must be emphasised that this is simply a first draft based on Stokes, English Vernacular Script, and will undoubtedly need considerable refining. In particular, I use the terminology that I am familiar with, but in a future post I plan to produce a table of equivalent terms. For now I am more interested in the principles, though: is this a valid model? Please post comments below!
Components of Letters and Punctuation
These are all the components that a letter must have to be recognisable as such (what Parkes calls ‘essential elements’):
a: bowl (also called bow or lobe), back
æ: a-component (comprising bowl), e-component (comprising eye, lower curve/bowl)
b: ascender, bowl
c: hook, bowl (or is this something different, e.g. a lower curve?)
d: bowl, ascender or back
e: eye (comprising hook, tongue), bowl (or is this something different, e.g. a lower curve?)
f: hook, tongue, descender or downstroke
g: mid-component or bowl (see below), tail
h: ascender, arch (comprising upper curve and also down-stroke/minim: see below)
i: minim
k: ascender, upper branch, lower branch
l: ascender
m: minim, middle arch, final arch
n: minim, arch [For minuscule only, but majuscule n is sometimes found in otherwise minuscule script]
o: bowl
p: bowl, descender
q: bowl, descender
r: Either minim and hook, or descender and arch (see below)
s: [See below]
t: top-stroke, bowl (or is this something different, e.g. a lower curve?)
thorn: bowl, ascender, descender
eth: bowl, back, cross-stroke
u: minims
wynn: descender, bowl
x: right-to-left stroke (comprising north-west and south-east branches), north-east branch, south-west branch
y: [See below]
z: top-stroke, diagonal stroke, bottom stroke (optional tail)
7: top stroke, descender (but is this descender the same as other descenders?)
punctus: point
punctus versus: point, comma
punctus elevatus: point, up-stroke
punctus interrogativus: point, top-stroke
Components of Allographs
These terms can undoubtedly be adjusted and standardised more; I put them here as a starting-point. Even here, though, there are problems, particularly with s and y, but also with f, g and r. This is because these letters can have drastically different forms, even in the Anglo-Saxon period, and so one cannot easily find components which are necessarily common to each. To resolve this we need to go to the level of allograph:
f Caroline: hook, tongue, down-stroke
f Insular: hook, tongue, minim
g Caroline: bowl, tail
g Insular: top-stroke, mid-component, g-tail
r Caroline: hook, down-stroke
r Insular: hook, descender
s round: upper curve, lower curve
s tall: hook, down-stroke
s Caroline: hook, down-stroke
s long: hook, down-stroke, descender
s low: hook, descender
y straight-limbed: upper left branch, right-to-left stroke (comprising upper right branch, y-tail)
y round: top-to-bottom stroke (comprising upper left branch, y-tail), upper right branch
y f-shaped: upper branch, lower branch, y-tail
Other allographs include the remaining distinctive letters of Caroline script, plus some others:
a Caroline: Also has a head
c horned: Also has a horn
e horned: Also has a horn
d Caroline: Has an ascender (not a back)
d Insular: Has a back (not an ascender)
You might reasonably ask why Caroline letter-forms are included in a description of Vernacular minuscule: the answer is simply that these forms appear often in otherwise Vernacular script and so need to be taken into account.
Sub-Components
Some components have components themselves:
bowl: south-west quadrant
eye: hook, tongue
ascenders and minims: top-decoration
descender and minim: foot
back (of eth and insular d): tip
back (of a and æ): foot
arch: shoulder, downstroke, foot (I argue that the downstroke is not a minim for my per
The ‘arch’ is problematic but refers to the common part of h, m, n, and Insular r, including not just the top curve itself but also the full down-stroke and foot. In Vernacular minuscule the down-strokes here are not necessarily treated in the same way as minims and should not be considered as such (unlike Gothic Textura, for example).
General Features
These features can apply to any letter and (in principle) any stroke:
Pen: Thick/thin/medium width; blobby; flat/angled
Writing Angle: Upright, forward-leaning, backward-leaning
Rotundity: Angular, rounded
General Aspect: Messy, neat; narrow; spidery; bulging; heavy; shaded; level cue-height
Connectedness: Separate/conjoined/ligatured
Component features
These are features of specific components or sub-components:
Ascenders: long/short/minim-length; sway-backed; backward-leaning
Back (of eth and Insular d): Bilinear/45°/upright; long/short; straight/concave up/concave down/broken; backward-reaching
Back (of a and æ): Long, straight/round
Bowl: Concave left/concave right/angled south-west quadrant/round/square; horned; point-topped/flat-topped
Descenders: Turned left/curved/straight/tapering; foot/no foot; long/short/minim-length; angled forward/vertical/angled back (esp. relevant for 7)
Eye: Squinting/open
Foot: Angled up/angled down/horizontal/absent; long/short
Hook: Angled/flat/bulging/high; long/short; deeply split; looped
Mid-component (of Insular g): Angular; large/small; Hangs from left/middle/right
Minim curve: Shoulder angled/bulging/rounded; deeply split; downstroke rounded/straight; downstroke turned in/out
Tail (of g): On left/middle/right; narrow/wide/oblong/round/angular; open/closed/3/4 closed; tip horizontal/curved down/curved up
Tail (of y): Hooked/straight/curved
Cross-stroke (of eth): High; concave up/down; hooked down/up/no hook; through/not through
Tip (of back): Vertical/turned back/turned down/wedged/trailing left
Top treatment (of minims and ascenders): Vertically symmetric, horizontally symmetric, wedged, barbed, clubbed, approach-stroke, back-reaching, flat-topped, split, forked, tapering on left, blobby, none
Top Stroke: Convex/flat/~-shaped; left hooked up/down, rising; at cue-height/above cue-height; short/long; turned-up right
Tongue: Angled/horizontal/convex; long/short; turned down/turned up; rising; thick/thin; on left
Letter features
These are specific to particular letters. Arguably some (or even most) of these are allographs, but the distinction is a difficult one:
a: Caroline/semi-Caroline/cc/teardrop-shaped/round/square
æ: As for a and e
d: [Arguably round/upright (or Insular/Caroline), but these are probably best covered as allographs and via the back component.]
f: Caroline/Insular
g: 3-shaped/S-shaped (or make these features of the mid-component?)/Caroline
h: Caroline/Insular
k: c-shaped/arm up; descender
l: Round bottom/angled bottom
n: Majuscule/minuscule
r: R-shaped; Wulfstanian
s: Tall (or long) before t; tall (or long) before wynn; tall (or long) initial; tall (or long) with t only; tall (or long) in ligature with t only; tall high wedge; high round
t: Turned-up/down toe; closed
thorn: Predominant/initial/conventional
eth: Flagpole, predominant/initial/conventional
x: Bilinear/long southwest quadrant; three-stroke
y: Bilinear; dotted/undotted; high right (straight or round only); hooked right (straight or round only)
z: Bilinear, high, long southwest stroke; diagonal descender
7: High top right
punctus: On baseline/at mid-height/at cue-height
Again, there will undoubtedly be additions and changes to this list, but it should provide a useful starting-point. In the next post I hope to take this further by applying it to some more concrete examples.
Hi Peter,
Where does ‘hook’ come from? Is it the same in ? Alex Rumble (he was London trained) taught me ‘shoulder’, which I like a lot, and which was used for long, low and high s, and . I’ve never come across ‘eye’ in e before; I was taught ‘bowl’. Anyway, those might be palaeographically dialectal, but the terms you use for are slippery: I don’t get ‘topstroke’, which is ambiguous and could be the little stem above the ‘cross-stroke’ (the term for the horizontal stroke). I think eth has a cross-stroke, and a stem and bowl. And with the nota, plus some other graphs, I wouldn’t say ‘descender’, since often there’s no descent, but ‘downstroke’. Ah well. Thanks so much for this. So thought-provoking. Alex has that great piece on A-S Mss in Richards (is it?) and in the collection he himself did. All his terminology is there.
ET
Many thanks for the feedback, Elaine.
Like most people, I’m mainly following the terminology I was taught, plus some inventions in cases where I’ve not seen anyone describe the things I want. As I’ve written in the text, this is the list of what I’m familiar with as a starting-point for discussion, not what we plan to continue using in the future.
Regarding e, if we call the ‘eye’ a ‘bowl’ then how do we distinguish the bowl of, say eth? It’s clearly a very different thing, produced in a different way. Ker sometimes calls it the ‘2-like stroke’ (Catalogue, p. xxix), but that’s clearly applicable to only one particular form.
Point taken for the t, although the cross-stroke of t is clearly different from that of eth, I think, and so needs to be distinguished (‘t-cross-stroke’?). Part of my point is that everything called a ‘cross-stroke’ should be produced in the same way, so that we can search for common aspects in the way the letters were written (what I’ve rather clumsily called ‘style’), not just in the shape of particular letters.
Point taken for the nota, although the question again is whether, for those cases where the downstroke does descend, if it’s constructed in the same way as a ‘true’ descender or not. Again, the downstroke (or whatever) of the nota is constructed very differently from the downstroke of, say, r; the latter may be a minim or full descender but can also be something in between. Perhaps the answer is that the nota (and some other graphs) can have a downstroke, or a minim, or a descender.
I’m familiar with Alex’s terminology, and I am planning to collate some different ‘dialects’. It’s not always clear what he means, though: in Richards, for example, he lists the terms but doesn’t define any of them.
I am surprised that Lindsay’s classic article in Palaeographia Latina has no place here. Parkes provided a glossary which makes an important distinction between essential and subsidiary elements. His term arches seems satisfactory. His concept of a personal idiom is worth more attention.
I am sorry that serifs have no place in your vocabulary.
I can see no advantage in ‘southwest’
Crucial is the dialogue with calligraphers which establishes whether terminology is descriptive or analytical. Better, surely, to describe real pen movements. And better still to describe English scripts in a standard terminology than to create new terms for minor regional variants. Surely you need to be talking to real palaeographers like Marc Smith?
Many thanks for your feedback, too, David: again, it’s much appreciated.
I’m very conscious that I still haven’t brought in much from other scholars yet, despite your urging: as I mentioned at the top, the idea of this was simply to list what I’ve been using in the past as a starting-point, to then collate others’ systems as well.
Parkes’s terms are very satisfactory for what he was doing, I agree, and for most purposes, but I don’t think they go to the level of detail that I want here. The point of this is to have a level of precision which allows one to find examples of particular features in a database of 1200 scribal hands: that’s obviously very different from the normal needs of palaeographers.
The essential/subsidiary distinction is certainly important, but that’s to come in later, I think: at this point I’m just trying to provide a list for debate. And his concept of personal idiom is exactly what I’m trying to get at here, though with different terms. The problem is that the computer requires a much higher level of precision and absence of ambiguity than we’re used to, so while Parkes’s work is excellent for normal purposes, I think it needs a bit of extending for these purposes.
Point taken regarding serifs, and I’m happy to use them. My problem is that ‘serif’ tends to be conceived as both top and bottom of minims, but of course in Insular and Anglo-Saxon script the two are very different and so need to be distinguished for the computer. I agree that ‘top-serifs’ is much more elegant than ‘top decoration’, but I wonder if the deeply split ascenders of mid-eleventh century vernacular script can reasonably be called serifs. I have no problem at all with ‘foot-serif’, except that it again raises the question whether it’s the same as the serif on a descender, so do we then need to distinguish between ‘minim-foot-serif’ and ‘descender-serif’?
I also dislike ‘southwest’ intensely, but after eight years of looking have not found an alternative. Please let me know if you have one!
I agree with the crucial distinction between ‘descriptive’ and ‘analytical’, and I certainly do have discussions planned with calligraphers. Please encourage any to contribute to this page, too. Indeed, as I wrote in my reply to Elaine, that’s very much part of the reason for separating out components in this way: to try to identify those that are produced in the same way across different letters and to compare them.
If there is such thing as a standard terminology, and if it has the detail and precision that is necessary for this project, then I’d be extremely happy to use it. I haven’t found one yet, though, and my understanding from talking to Marc and others is that the Comité hasn’t either. I wonder, too, about your point of minor regional variants: isn’t part of the problem precisely that different regional scripts are distinctive in different ways and so need to be described with different terminologies that focus on different aspects? For the purpose of this project, if we can collate the main terms used for particular components, then we might at least be able to provide a thesaurus so that you could search with Lindsay’s terms, Elaine could use Alex Rumble’s, and I could use my own. I’m still sceptical that this will work, since the terms don’t usually equate with each other one-to-one, but it’s worth a try to see what happens.
Fascinating discussion. Thanks! It is perhaps worth pointing out that it is our intention to provide a fully *illustrated* glossary of the terms that we use in DigiPal. So, if we say “south-west single-footed up-flick”, there’ll be a manuscript image to show exactly what we mean by that. The hope is that this project will be very different to the situation in print where you have limited plates and have to try and guess exactly what a palaeographer might mean (and hope that he/she is consistent in their usage of terminology). Which is not to say that we shouldn’t work with existing terminology as much as possible, of course. Early days
I’m a little worried by the way that ALLOGRAPH, IDIOGRAPH, and GRAPH all have a relation to GENERALFEATURE in the current model.
Certainly I think it is acceptable for a model to represent the overall layout of the domain and then implement specific “business rules” in processing stages. However, I wonder if this might be better kept at the level of *horizontal* distinctions, ie. applying rules to allow for different treatments of members of the same class. For example, I think it’s okay that the model just has a many_to_many relation between allographs and components, and then it’s up to the “business rules” to state that ALLOGRAPH ‘lowercase d’ can NOT have a descender.
However, because allographs, ideographs, and graphs currently all have a many_to_many relation with GeneralFeatures, you are forcing the “business rules” to make *vertical* distinctions as well. For example, I’m not sure that General Aspect should be a feature applicable to ALLOGRAPHs, but the way the model is at the moment it is the business rules that have to make the distinction between ALLOGRAPH and GRAPH in this regard. In other words, there are surely features that *only* a GRAPH (and possibly IDIOGRAPH) can have, but currently this can only be captured through the business rules, not the model.
This post is referred to on the following web page: AAA – ΑΔΛ : actualité paléographique et ontologie des formes alphabétiques « Paléographie médiévale
This post is referred to on the following web page: Modélisation des signes graphiques (1) « Paléographie médiévale