opération manta

manuscript

theology

manuscript

On this page:

A writing system I use to explore what I find interesting about manuscripts. For now, I’ve tried to make the system 1] a bit more phonetic in its spelling than standard English, and 2] fun to write in cursive!

It’s also an abjad, because abjads are cool (and potentially efficient, see spelling).

It is a work in progress.

There is a github repository associated with this project where I plan to put font files and the like.

consonants

29 English consonant sounds are represented. Rather than simply come up with 29 unique glyphs, we’ll devise a system to organize these sounds in way that makes writing them more manageable.

consonant sounds

1. Sounds are organized into three families based on the general area where they are pronounced in the mouth (see rows in the table below).

2. Within these families, sounds are organized into groups based on three factors of similarity (see columns in the table below):

  • Unvoiced and voiced variation on the same sound (e.g. t/d)
  • Related pronunciation (e.g. both t/d pair and θ/đ pair are pronounced with the tip of the tongue at the front of the mouth)
  • Reoccurrence of a sound within compounds (t/d contained in tʃ/dʒ)
Consonant sounds a. base letter b. related c. compound
unvoiced voiced unvoiced voiced unvoiced voiced
1. lips bilabial p b
m w cw gw
labiodental f v
2. ridge alveolar t d θ đ tʃ dʒ
s z ʃ ʒ
n
l
3. back palatal r y
velar k g ks gz
ŋ
guttural h ʔ

This way of organizing our consonants will determine how we write them. Instead of having 29 random glyphs, we’ll have :

  • three families of glyphs (lips consonants, ridge consonants, and back consonants)…
  • made up of twelve unique letters (the twelve rows)…
  • with a maximum of six variations of each letter (the six columns).

Let’s see how this plays out graphically.

writing consonants

Each consonant follows one of three shape patterns, and each shape pattern corresponds to one of the three pronunciation families on the consonant sounds chart above. In other words, a consonant letter’s general shape — specifically, if it has a stem and where the stem is located — indicates approximately where a letter is pronounced in the mouth.

I write 29 consonant sounds with only twelve letters, one letter for each row of the consonant sounds chart above. Diacritics are then used to distinguish sounds in the group (columns on the chart below).

Consonant
patterns 		a. base b. related c. compound
unvoiced voiced unvoiced voiced unvoiced voiced
1. lips ! ; , . ' : "
2. ridge ?
3. back ;

Example:

Here’s how the above patterns play out with the letter p:

  • p has a stem on the left side of the letter, meaning it is a sound pronounced at the lips. It represents the p sound.
  • If you take the letter p and add a voice mark on top of it, you get b. This represents the b sound (p + voice).

The 29 written consonant sounds follow the above patterns. One exception for practicality: it is not necessary to write the voiced diacritic for voiced consonants with no unvoiced counterpart (m, w, n, l, r, ŋ).

Here’s what that looks like:

Writing
consonants		 a. base letter b. related c. compound
unvoiced voiced unvoiced voiced unvoiced voiced
1. lips bilabial p (p) b (b)
m (m) w (w) q (kw) W (gw)
labiodental f (f) v (v)
2. ridge alveolar t (t) d (d) F (θ) V (đ) T (tʃ) D (dʒ)
s (s) z (z) j (ʃ) J (ʒ)
n (n)
l (l)
3. back palatal r (r) y (y)
velar k (k) g (g) x (ks) G (gz)
N (ŋ)
guttural h (h) H (ʔ)

That’s it for consonants.

vowels

Vowels are still evolving in how I write them. This flexibility doesn’t tend to impact the rest of the writing system, however, since vowels are represented as diacritics that freely float below the consonant that precedes them.

For now, I tend to be able to get away with writing only 15 vowel sounds, which, much like consonant sounds, are organized into families (see rows in chart below) so that we don’t have to write 15 unique glyphs.

Since vowel pronunciation is so heavily influenced by dialect, I have provided example words alongside of IPA. The latter is only accurate for my accent.

vowels short long diphthong other
a nap (æ) cape (eɪ) aha! (ɑ)
e pen (ɛ) peek (i)
i nib (ɪ) tide (aɪ)
o hot (ɑ) note (oʊ) loud (aʊ) raw (ɔ)
u mud (ʌ) muse (ju)
oo room (u) book (ʊ)

cursive manuscript

connecting letters

Writing this script in cursive is not as straightforward as typical English cursive with a latin alphabet because letters do not consistently attach at the writing line.

Instead, how a letter is written depends on two factors:

  • Whether the letter is at the beginning, middle, or end of a word.
  • Whether the stroke of the pen approaches this letter from the top, middle, or bottom.

The table below summarizes how letters change shape and how they attach based on these two factors. Letters not shown on the chart (k, N, h) don’t change.

cursive beginning middle end
top middle bottom top middle bottom
p
m
f
t
s
n
l
r

Since cursive is partly a matter of ergonomics, partly a matter of esthetics, there is flexibility here. For example, I often find that I like to use the middle form of letters in the t family at the beginning of one-syllable words (see examples in next section).

connecting words

Words aren’t always written from left to right with spaces in between; they can succeed one another in other ways as well.

For example, the most obvious way to write the words “the pen” would be in the following way:

However, I would probably write it like this:

In addition to choosing the middle form of the letter t, I have also extended the tail of this letter in a way that causes the second word to “sit” on the first.

We could also write the words “this pen” like this:

We can briefly explain how words overlap like this in three points:

  • A word can sit on a horizontal stroke belonging to a previous word if that horizontal stroke is at the writing line.
  • The two words remain distinct in that aren’t connected with a continuous cursive stroke.
  • Word order remains discernible in that the first letter of the first word precedes the first letter of the second word.

As long as word order remains clear thanks to these constants, there’s room for quite a bit of flexibility.

spelling

vowels

An advantage to using diacritics for vowels is that you can omit them when they’re not strictly necessary without changing the overall look of a word. Vowels are, in fact, optional.

Now, unlike in semitic languages, we can’t omit vowel information entirely and expect English to remain intelligible. Still, a good number of vowels can in fact be left out.

To demonstrate, take the consonants below:

Without vowels, this spells out “W’TH ’M’J’S ’N’T T’P’L’J”, which is not very readable, apart perhaps from the first word, “with”.

First, we can use a tie (like a tie in music) to indicate when two consonants follow directly, with no vowel between them:

In reality, such a tie is unnecessary in this case because the letters Hnt (or even hnt) will always be the word “and” unless otherwise specified. Frequent words like this don’t need as much detail.

Without vowels, the second and fourth words are tougher. So we start by writing in any vowels that are in stressed syllables:

This gives us “WITH IM’J’S AND T’POL’J” (stressed vowels underlined). The second word is now clearly “images” even though two vowels are missing.

It turns out that the final word ends in a vowel, but this isn’t obvious given the way we’ve written it. It’s generally good to write final vowels. Since this word ends in a long-e sound, and because long-e endings are so frequent, they’ve developed their own easy ending: a lengthened letter tail and a long parallel line placed beneath it.

Even if we don’t write the final vowel, we should at the very least extend the tail on the final letter to indicate that a vowel of some kind is present.

The vertical mark placed under the p in the last word can be useful to specify which of several marked vowels is emphasized.

This gives us “WITH IMAGES AND T’POL’JEE”. The last word could be “typology” but it could also be “topology”. In the majority of cases, context makes choices like this obvious. Still, when an uncommon word appears for the first time in a text, it’s a good idea to write more vowels.

In the end, we might end up with something like this:

In this example, we’ve omitted two thirds of our vowels without sacrificing intelligibility. When this is developed as a reflex, writing becomes more efficient.

To summarize these rules of thumb:

  • Very common words (especially conjunctions and the like) barely need diacritics.
  • Ties can be used to link two consonants. Series of three consonants are linked with an upside-down tie (see sample text below).
  • Vowels should be placed under stressed syllables in words that aren’t immediately obvious without them.
  • Final vowels should be indicated with a tail.
  • When several vowels are provided in a word, the emphasized vowel can be marked.
  • Vowels and other markings can of course be discarded after the initial occurrence of a frequent word in a given body of text.

Ok, now we can see how it all comes together.

sample text

Transcription below.

Title: “Asterix in Corsica”

“For most people, Corsica is the homeland of an emperor who has left pages in history as indelible as those inspired by our old friend Julius Caesar. It is the country of vendettas, siestas, complicated political games, strong cheese, wild pigs, chestnuts, succulent blackbirds and spry old men watching the world go by…”

fonts

As much as I’d like to make one, currently there is no cursive font for this script. But there is a bare bones, rough-around-the-edges print font which I made for the purpose of documentation, and which can be downloaded from the github repository. This font only includes consonants.

The chart below summarizes how inputs are mapped: