Creative Arabic Calligraphy: Kashida, Ta’jîm and Tashkîl

In this lesson we’ll look at two more sets of “design elements” before we start putting it all together to create letterforms and compositions.

Kashida: The Art of Justification

Spacing, which we have studied in an earlier lesson, should not be confused with kashida (aka tamdîd, tatwîl or madd, all of which means “elongation” or “stretch”), the equivalent of justification for Latin-script texts. To justify a text is to make sure all the lines in a block of text begin and end on the same level. When a block of text is not justified, it is ragged left or right, like this:

In contrast, below is text that has been justified. In the Latin script, this is done by expanding or contracting the space between letters. In the Arabic script, it is done by stretching the letters themselves, and this is not necessarily in the context of equalizing a text: kashida can be applied to a standalone word just to make it more interesting.

I want to make a subtle distinction here between stretching actual letters, and extending the connecting stroke to the next letter. For our purposes, we’ll refer to the first case as stretch or mashq, and to the second as lengthening or tatwîl. Kashida will refer to justification in general as achieved by either or both. 

The text above shows examples of lengthening only, as a computer font can do no more. Lengthening is only applicable to some connecting letters, when they do connect (that is, when they are in their initial or medial form):

Lengthening is not applicable to boxed letters, or to be more accurate, is unwanted there. This is because boxed letters, which are strongly horizontal, can be stretched in their body —at least in rectilinear scripts. Not so in round scripts, so stretching is largely exclusive to Kufic.

Stretching is preferred over lengthening whenever possible, because it doesn’t create a big white space in the texture of the text. Therefore, if a letter can stretch, to lengthen it makes no sense, and looks wrong. Dâl, for one, is not a connecting letter, so it could not be lengthened anyway. Boxed letters are useful because they can stretch even when they are final or isolated.

When a letter ends in a flat tail, that can be lengthened as well:

This leaves us with a few letters that can neither stretch nor lengthen in the positions shown:

Inevitably, some words are going to be impossible to stretch.

Here’s a list of words that offer zero to many possibilities for justification. A red cross means that option is not possible; an orange one, that it is possible, but not desirable. Let us examine why.

In the case of ??? lengthened to ???????????, we have already seen that when a letter can stretch, it makes no sense to lengthen it instead. 

The next word has two lengthening options because it contains two letters that can be lengthened. What difference does it make which we use? The difference is beauty. While the ????????? option is not wrong, it is, simply put, a boring option, because the long stroke is bang in the middle. The word is then divided into two somewhat equal parts, which appears static yet not accurate enough to be satisfying. 

In contrast, the uneven split in ?????????? is dynamic—the final Ayn looks much better on its own, if split there must be. There is also the fact that the Yâ’ here is the long sound, and matching the lengthening of the line to that of the sound is both pleasing and artful.

The final word offers several possibilities and it’s not so easy to discard some. ????????????? was discarded because, again, we have a box letter being lengthened instead of stretched. It is even more undesirable here because the two verticals of the Tâ’ and Lâm have been split apart, and verticals, as we saw in our lesson on Spacing, are always better close together. 

The next discard is based on sheer unattractiveness; it is not necessarily wrong to have more than one kashida in the same word, but it needs to look good and balanced, and this is not the case. Perhaps if they were both the same length, the result would be less half-baked, and perhaps context would make the difference. It’s not cut and dried. 

The same goes for the other options—they may or may not work, depending on context. Personally, I would choose the stretched option if I wanted the word to remain blocky, or one of those that isolate a single letter ( ????????????? or ???????????? ), for the dynamism of the result.

Historically, kashida was used for various reasons, some aesthetic and some functional. The list offers suggestions for introducing this device in a composition to great effect.

Aesthetic Use

• Justification itself is the most obvious use: leveling the beginning and end of every line in a paragraph so that it is altogether shaped like a regular block, or adjusting the length of a single line.
• Introducing breathing spaces in a body of text, and/or a pleasing rhythmic pattern of spaces.

• Positioning letters (pushing them out of the way, so to speak) so that ascenders and descenders don’t clash,
  especially when lines are quite close together, but also to create alignments within the text.

Functional Use

• A word is often extended (to the point that it takes up an entire line) to mark the beginning or end of a paragraph. This was particularly common when Arabic used no other punctuation.
• Bringing attention to important words in a text, the equivalent of bolding or highlighting them.

Other considerations

• Where possible, and if not overridden by other considerations, kashida is made to match a long sound as mentioned above.
• Two-letter words are generally not stretched or lengthened.
• Avoid “stairs”, or the superposition of elongations on two consecutive lines, unless that is done by design and well studied.

Exercise 1

Here is a sentence with two stretchable letters and plenty of connections that can be lengthened.

(This is an old Lebanese proverb that says: “In April nature becomes like a bride.”)

Print out Exercise1.pdf which is included in the downloads folder. You will find a grid over three lines, with the first and last letter of the sentence already in place. Your task is to distribute and justify these four words over the three lines in a way that looks nice to you. 

You can try several possibilities; I recommend starting by just trying out the kashidas without worrying about aesthetics. The restrictions in place actually make it easier by not allowing infinite possibilities. Exercise1solutions.pdf shows two possibilities and points out what makes them interesting.

The letters are very thin so that you can copy them by hand, preferably freehand even if the lines don’t come out straight. It’s quicker so you’ll be able to focus on the exercise rather than on handling the ruler. Also, it’s about time we start drawing letters, even skinny ones, to warm up for the next lesson!

Ta’jîm and Other Diacritics

These are a number of additional signs that can accompany an Arabic text to indicate sounds not included in the pure alphabet. They are often neglected in the rectilinear scripts, which predate ta’jîm, and much more prominent (compulsory, even) in the round scripts, but it is good to have them in our palette, so to speak, as potential for added layers of visual interest. 

However, we must bear in mind that there is a certain hierarchy to be respected in their usage. Vowel marks, for instance, cannot be included if pointing is not. The diacritics are discussed in order of precedence, starting with the highest rung of the hierarchy, unless specified.

As this is not a language lesson, I’ll only give minimal info regarding the sound and function of these marks, focusing instead on their shape and position, i.e. what we need to know to be able to create with them.

Ta’jîm

These are the diacritic “dots”, or for a better term, the pointing used to differentiate letters that share a letterform, such as ? and ?. Ta’jîm, or i’jâm, literally means “foreignization”, because, like the other sets of signs that make reading easier, pointing was introduced for the sake of populations that were not native Arabic speakers and therefore would struggle with the correct reading of the Qur’ân.

The dots are at the top of the diacritic ladder of hierarchy:
they need to be present before any other level is included. If they are
omitted, then the script should be entirely bare, save for the isolated
hamza (see below) which has a semi-letter status.

Tashkîl

Literally “forming [elements]”, these are phonetic guides or diacritics.

Hamza:

Shaped like a small Ayn without returning tail. Depending on the sound accompanying it, it
can be placed over the Alif, Wâw and Yâ’ (in which case the Yâ’ loses
its dots), under the Alif, or independently on the baseline. The size of
the hamza changes subtly: when attached to one of the letters above, it
tends to be tiny, like other diacritics. When it’s isolated on the
baseline, it is more sizeable, without attaining the full size of a
letter. Scaling the isolated hamza is an interesting design conundrum.

Shadda (“stress”):

This sign is the equivalent of doubling a consonant. It has the
shape of a tiny Sîn without a bowl (somewhat like a mini-W) and can be
found above any letter except the Alif. Another aspect of the shadda
is that it pairs up with the short vowels as shown below: fatha and
damma over it, kasra under it (not under the baseline, exceptionally).

Hierarchy-wise, the shadda and diacritic hamza are in close contest.
However, the isolated hamza is a semi-letter and as such, takes
precedence even over ta’jîm.

Harakât (“movements”):

They indicate the short vowel sounds that the alphabet lacks. Fluent speakers are expected not to need them, and even today they are mostly used in learning texts, children’s literature, religious texts where pronunciation is vital, or
here and there where a word is ambiguous. And, of course, for purely aesthetic purposes. These signs are:

• Fatha: A diagonal stroke from upper right to lower left, or
  horizontal—never vertical so as not to be confused with the dagger
  alif. Always above the letter.
• Kasra: Same shape as the fatha, but positioned below the letter.
• Damma: Shaped like a tiny Wâw, always above the letter.
• Sukûn: A hollow circle indicating “stillness”, an absence of sound.

• Madda: Shaped like a tilde, or possibly a long horizontal dash, but it should not be able to be confused with the fatha. It only appears above an Alif, doubling it; so in effect it is the Alif’s own shadda.
• Dagger alif ??? ??????: This is an archaic way of writing the long a
  sound, as the Alif was originally often unwritten within a word; that’s
  why very old words, including the word Allah itself, have Alif vocalized
  but not written. As a result, the dagger alif occurs frequently in the Qur’an, but rarely
  in daily writing where words are spelled in the modern way—except words
  so old and common their spelling has been in continuous use
  unchallenged, such as “this”, hâdhâ ???, but even there, the pronunciation is assumed to be commonly known, and the sign is not used. It is a short vertical stroke
  above and between two letters, where the Alif sound should be.
• Wasla: A “joining” symbol written above the Alif to suppress its
  pronunciation, joining two words without a glottal stop. It looks like a
  tiny Sâd without a bowl, and is only ever found on the first Alif of a word or of the article.

Evolution and Diversity

All of these symbols underwent a long evolution before settling in the form
that we use today, and their entire history can be followed within Kufic
texts, from complete absence, to early experimental forms, until at
last in the so-called Eastern Kufic, they are seen fully mature. Today’s average Arabic reader, looking at the text below, would
automatically attempt to recognize the letters based on the red dots
(until they realized that yielded no sense).

But

these dots are not pointing at all! In this early vocalization
system, they indicate the short vowel sounds (harakât, described below). The actual pointing
is done in small dashes close to the letters, which themselves look
exactly like what we now use for vowel sounds, so things were altogether
turned around in the course of their evolution. 

The example below, though later in time, has no diacritics at all other than pointing, and even that only where a word is too ambiguous. Thus the word ??????, near bottom right, has a full set of dots, while ??????  just above it, was judged too unmistakeable to warrant them.

Contemporary to this is another manuscript where the system, in contrast, makes elaborate use of colour:
red dots are still used for the short vowels, hamza are added in green,
shadda in gold. The pointing is now even more discreet as the dashes are
the same colour as the body text. You can just see an example on the
bottom line, towards the left, where a thin dash on the Nûn of ?? is
nearly eclipsed by a large red dot.

One
needs to also be aware of regional differences, which continue to this
day—perhaps not in typed media, but at least in handwriting and
calligraphy, particularly religious. 

In Egypt, for instance, the final
Yâ’ may be unpointed, causing the unaware reader to confuse it with a
“broken alif”, written like an unpointed Yâ’ but pronounced like an
Alif. North Africa, the province of the Maghribi scripts, followed its
own line of evolution and so held on to some bits of early diacritics.
In particular, the Qâf and Fâ’ are pointed differently than they are in
the Mashriq: in the text below, the letter circled in red, looped with a single dot above, is a Qâf. The Fâ’ is circled in blue, and its dot is below the loop.

Notice also another use of colour: the vowel marks, including a dagger alif, are the same colour as the body (and they are horizontal dashes, a signature of Maghribi), except for the sukûn. The latter is blue, like the shadda, presumably because they are both silent. There is another sign very much like sukûn, except for its green colour, over some Alifs: that is a wasla. One more circle, orange and full, stands for the hamza.

In the elegant Eastern Kufic page below, colours are used again, but more sparingly and within a fully-formed system. 

Pointing is in large round dots, as if the previous ratio of bold lines to tiny dashes had been reversed, but is actually the most subtle part of the diacritics, as the dots are gold with only a fine outline. 

Again the vocalized signs are all in one colour (red), while those that are mute, like the sukûn, are blue. 

There is an additional detail of interest: circled in white, you’ll notice tiny letters, ? and ?, under their in-text equivalent. When pointing was still a novelty, it became customary for a time to annotate non-pointed letters with their tiny isolated self, as done here, for clarity. 

Getting Creative With Pointing

The number and basic position of the dots, which identify letters, have been described in Anatomy of the Letterforms, and there’s little more to be said in the way of rules. There is much, however, to be said about what can be done with pointing as an aesthetic element. Because they’re so simple, and because calligraphy is about beauty more than function, dots are a singularly free part of the script.

Shape

The shape of the “dots” is by no means necessarily round. In fact they seldom are, looking at historic examples. We have already seen them written as dashes, and in the rounded scripts they are canonized as the rhombus shape created by the reed, which in the geometry of Kufic can translate to a dynamic square (diamond). 

There is no restriction at all on their shape, in theory, except good taste: the more fanciful the shape, the more noticeable they are, and that can quickly become visually cloying. Even when working with a short sentence or a single word, it is always good to estimate the number of dots it contains before getting overly creative with them. 

Here are two examples from my own work where the shape of the dots was determined during the design process, to enhance the composition.

The word ????? only contains three dots, and this design kept their repetition small. In addition, the sobriety of the design needed to be balanced with intricate touches. So I chose an unusually detailed shape for the dots, which worked beautifully as the intricacy of the pointing sets off the simplicity of the rest and gives a sense of scale, and vice versa; plain round dots would have made the whole composition look plain.

This next word, ????, has double the pointing, and the design multiplies that by 18! Normally I would leave out the dots altogether, but in this case, the extremely rectilinear Kufic would have been too severe. So once again I balanced it out, but this time using the simplest of shapes—dashes. Being harmonious with the dominating rectangles, they don’t intrude on the composition but echo it on a smaller scale, and by the same token make it more delicate than it would have been without them.

Size

The size of the dots is equally free, but I recommend
not being arbitrary: Let the diameter be in a clear ratio to the line
width (1:1, 1:2, 1:3…). This isn’t only perceived as looking right; it also ensures that
you’ll always find a harmonious positioning. Otherwise aligning two or three dots can look very awkward. The same consideration applies to the distance between the dots and the line.

Positioning 

The two dots can also, for instance, be placed vertically rather than
horizontally. The three dots are usually in a triangle, but can just as
well be aligned. If the line width of the letters is substantial, as it often is in Kufic, a creative approach could be to place the dots within the line width, or overlapping it halfway.

To Sum It Up:

• Diacritics can be completely or partially omitted, bearing in mind the hierarchy of addition.
• Diacritics can serve as visual accents by taking a different colour from the letters (with the distribution of colour very much up to the calligrapher).
• There isn’t one final single system, even if most of those that were in use at one point or another are long out of use.
• Pointing in particular is highly flexible in shape and location.
• Diacritics can be modeled closely on the style of the letters, or made to completely contrast; for instance, freehand cursive diacritics to give movement to a very architectural Kufic.

Exercise 2

Below is a pointed and accented sentence. Print out a few copies of the unaccented version, included in the downloadable folder (Exercise2.pdf), and experiment with adding different levels of diacritics, using different colours in different ways, and modifying their shapes and placement to your taste.

(This is another old Lebanese proverb: “Roaring March is the month of earthquakes and downpours, with seven big snows, not counting the little ones.”)