How to Judge Italic Handwriting

By A. S. Osley
Date Added: 21/07/2009


Writing-masters, particularly the lesser fry, have traditionally advertised the uniqueness of their wares, often introducing into the accepted script idiosyncratic and pointless modifications for their greater glory. The tradition has been well preserved in the italic field since the war. This article, condensed from the original version in Visible Language (Vol XIII, No.1) identifies from the texts of the period the essential characteristics of classic Italic, and this enables you to judge how far a given model diverges from the norm.

The information is taken from seven 16th century writing-masters - five Italians, one Netherlander and one Spaniard. The authors and their books are:

We shall now examine what these authorities had to say about such basic elements as slope, pen angle, letter proportion, length of ascenders, spacing, etc. Although each topic is treated separately, they are all inter-related parts of a single system. If one is changed or neglected, the whole style of the script is affected.


Slope is essential to cursive handwriting. When we write rapidly, it tends to increase. Unless, however, it is curbed, the letters will sprawl and break down. Illegibility follows. If we have learned to write from a model that has only a slight, though distinct, slope, we stand a better chance of remaining legible when we write fast. This fundamental point was instinctively grasped by the early writing-masters, though they had some difficulty in describing it precisely, possibly because the concept is best expressed in mathematical terms. Thus Fanti says: 'You must see that the long downward lines and the bodies [of the letters] slope a little to the right' and Tagliente refers to 'a little slope'. Vicentino is more roundabout: 'See that the letters slope as follows. . . .' He then writes the tag Virtus omnibus rebus anteit profecto with an exaggerated slant and adds: 'But I don't want you to tilt them quite so much as that.' Palatino states: 'Note that the chancery letter should lean a little forward, because in this way it can be written more quickly: moreover, if it slopes in the opposite direction, it is ugly and slow to write.' Augustino and Yciar say much the same - Augustino: 'Make sure that the letters slope towards the right, because you can write and form your letters more readily and more rapidly’; Yciar: 'Note that the whole should lean forward a little: for in this way it looks more elegant and can be written more rapidly.'

Mercator as so often, hits the nail on the head. He provides his readers with the drawing of a letter y on a grid, from which it is possible to calculate a slope of 5 degrees. Upright letters, such as those used in printscript, can confuse children who are learning to write. Uprightness is neutral, uncommitted; it will lead to either backward or forward slope as the lower or upper parts of letters are pulled forward under the gravitational force of increased speed. An alphabet with a slight slope removes ambiguity and leaves the letter-forms sufficiently close to those of the printed word for those starting to read.

Pen angle

The tool for writing on which the masters concentrate is, the quill, though Erasmus, and no doubt other scholars, often used a reed. Metal pens were also used. The quill is cut to a bevelled, chiselled edge which normally is at an angle of 90° to the vertical axis of the pen, though a slight angle was sometimes allowed. Here again many writing masters found some difficulty in defining the angle at which the pen should be held in relation to the line of the writing. Tagliente, for example, states:

‘My view is that you should not hold the pen by the edge of the nib; not with its full width, but obliquely, i.e., so that the full width of the pen always remains at an angle.' Palatino's recommendation is that 'the pen should not be turned in the act of writing, but must be kept steady at a slight angle.' Yciar gives clearer guidance, though he obviously had some difficulty in expressing it and 'adopted the formulation of an earlier Spanish scholar, Vanegas. He writes: 'You must keep the pen firm in the hand without twisting or turning it between your fingers; it should always' remain in the same position with your arm resting on the table. . . The position of the pen on the paper, when the letter is situated on the same axis as our body, should be somewhat tilted, as if the two tongues of the nib were placed on a die in such a way that the upper tongue points to the top right-hand corner, and the lower to the bott9m left-hand corner.' Since all faces of dice are perfect squares, the angle is 45°. Once again

Mercator gets to the heart of the matter in a few words: ‘The pen should, invariably be held in such a position that its broadest stroke would join the opposite angles of a square.' In practice, an edged pen, held constantly at an angle of about 45°, produces a satisfying sense of direction, a purposeful forward motion, which, when once it becomes second nature, is helpful in mastering the inefficient, directionless instruments used today.

Letter proportions

The character of the italic hand depends greatly on the proportion given to the small letters such as a, n, and o and on the ovoid shape used with letters that have a bowl such as a, d, g, and q. The question of proportion is especially significant. Insofar as models reproduced from wooden blocks can be regarded as a basis for measurement, the ratio of width to height used by Tagliente and Arrighi is about 3:5. Mercator's manual contains a diagram from which it can be deduced

That the ratio which he recommends is 2:3. There is, thus very little difference between the masters. Palatiino, however, in 1540, followed by Augustino a quarter of a century later, advocated a 1:2 proportion. This was a fundamental change. It leaves much less room for diagonal strokes; they have to be made at a sharper angle, thus increasing their width relative to the horizontal and vertical strokes. There is considerable risk that the letters will be too sharp or angular, producing a zig-zag or saw-edge effect. It was precisely on this point that, when the crisis of the classic italic hand came in the decade that followed the publication of Palatino's book, its trenchant critic G. F. Cresci rested his argument.

The crisis was resolved in two ways. In Italy and gradually throughout Western Europe, a new style, which led ultimately to the copperplate script, supplanted the earlier italic hand; in Spain however, primarily under the Influence of Francisco Lucas, the 1:2 ratio was dropped and the former 2:3 ratio restored. This more rounded letter was the hallmark of the Spanish bastarda, as it was called; it had a long life in Spain and is the true successor of the italic hand despite its name.

To return now to our authorities, Fanti is not specific about proportion. 'The chancery letter a,' he says, 'is derived from a quadrilateral. You must first place the pen above the line at a height. which seems to you appropriate, to the size of the pen which you are using. Then move the pen to the left and parallel to the writing line. When the pen has moved to half the letter-height that you have selected, then descend forthwith to the line with a curving stroke, which should make a small bend, curving not too much, but almost imperceptibly. When the pen has touched the lower line, them move it up in a straight line so that it joins the point from which you started the first stroke. Then bring the pen straight down, to the line again and, turning back up again, give the letter a little of a certain movement, which is called the "dead line" because it is almost invisible.' It can be seen that the width of the letter, which is made up of the initial horizontal stroke, plus the space necessary to make the curving downstroke, must be more than half the height of the letter.

Tagliente states, 'All the letters of the chancery alphabet are derived from the following oblong [he gives a sketch consisting of four dots].' He does not attempt to describe the ovoid body of the round letters, but again provides a little sketch, saying 'You should learn first the following body-shape, which is derived from a sloping oblong. To make a, one places a vertical line to the right of the body-shape.'

Vicentino does not analyse the construction of the letters but, like Tagliente, relies on a sketch. He observes that' All the bodies which rest on the line on which you are writing should be made so that they fall within an oblong quadrilateral, not a square [sketch] for, to my eye, the letter needs to be based on an ellipse rather than a circle. You will find that it will turn out to be circular if you shape it inside a square, and not oblong.' As indicated earlier, the ratio which appears to underlie the sketches drawn by Tagliente is about 3:5. In practice Vicentino writes a letter which is slightly narrower than that favoured by Tagliente.

Palatino's evidence is crucial. He writes: 'Chancery letters which have a body need to be half as wide as they are high, so that they make a double square: for if you make them within a single square, they will - so far as the proportions of their bodies are concerned, belong to the mercantile, not the chancery script.' Then, as though stepping back from the abyss he continues, 'But I don't say that it is necessary to observe these proportions each time that you have to write because that would be such a difficult and tedious business. I resolved, however, to set out this proportion, like the others which I have mentioned, for the benefit of those who wish to master every aspect of this art, both in theory and in practice. To make the letter a, you should begin with a horizontal stroke and with a light turn, go down with your pen, making the downstroke. Then with a diagonal turn, move up to join the horizontal and then come down again with another downstroke. . . Give the letter its proper roundness andelegance.' It is significant that Palatino, who had considerable influence in the 1540'S, changed his hand after Cresci had criticised its angularity, and brought it very close to the model that Cresci was advocating.

Augustino follows Palatino, but without any saving clause. 'From the point of view of geometry,' he says, the true chancery letter is based on two squares, not one, if you wish to shape and write it correctly.'

Yciar's testimony is ambiguous. He starts by saying that 'the chancery letter, when given its correct construction and dimension, should keep to the proportion and shape of a rectangle, the height of which is almost twice its width', but goes on to draw a pair of parallel lines, commenting 'the body of the chancery letter will be bounded by these lines so far as its height is concerned; its breadth, however, . . . will be equal to half the distance between these two lines'. He repeats Palatino's reservation: 'I do not mean that we must always observe this measurement in the chancery hand, but that, once we have understood the ideal proportion, we can use it as a guide and keep it as far as possible, especially when we are starting to learn it.' His examples of handwriting exhibit letters somewhat more rounded than those proposed by Palatino.

There is no ambiguity about Mercator. He gives an explicit diagram from which it is plain that he favoured a 2:3 ratio. In his text, he writes, ‘All letters that fall within the writing lines must have the same width as letter y, with the exception of m.'

Ascenders and Descenders

These are the lines which go above the band of writing (as in the case of h) or below it (as with g). The elegance of a script will depend much upon their length in relation to the other letters. Fanti recommends that 'the distance between one line and another in the chancery script should be four times the height of the letter (according to the size that you intend to make); and the ascenders and descenders should be equal in length' and should occupy these quarters of the space between the lines more or less'. Another way of putting it might be that the distance between lines is four x-heights: the band of writing occupies one x-height, the ascenders another, and the descenders a third. Three of the four x-heights are thus accounted for, the fourth being divided so that half an, x-height separates ascenders from descenders of the line above and the other half is the space between descenders and ascenders coming up from the line below.  Tagliente says, ‘All ascenders and descenders should be of equal length'; Vicentino: 'Take care that. . .all the long ascenders are of equal height . . '. similarly the descenders should be of equal length"; and Palatino: 'The height of all long down-strokes (ie, ascenders and descenders) should be  twice that of the body of the letter: they should be equal in length whether they go above or below the line’. Augustino has a sketch of four parallel lines with the small letters falling between the inner pair, ascenders rising to touch the top one and descenders going down to the bottom one.; later he states: 'The vertica1 lines which extend above and below the bodies of the letters should be regularly constructed to, the same height of or depth. Yciar’s opinion is that 'All the ascenders and descenders of the long letters, whether above or below the writing line, must be of equal size, their length being measured as the sum of their body height plus that of the band of writing.' Logical as ever, Mercator starts his book by asking his pupil to mark off with dividers equidistant points corresponding to the number of lines to be written and, then, having narrowed the dividers to the height of his writing, to construct pairs of parallel lines within which the written material is accommodated. Further on, he says that all the short downstrokes (such as those of i, m, and u) should slope equally to the right and be equal in length within the parallels. 'Of the remainder', those with ascenders rising above the parallels and those with descenders are twice as long, while those that have both ascenders and descenders are three times as long.' There is thus general agreement about the length of ascenders and descenders for handwriting.[2]

Many writing manuals today teach rather stubby models with ascenders and descenders of about half the relative length found in the chancery hand. Since, in writing cursively, we tend to shorten long lines, it is wise to teach the classic proportions.[3] Moreover, it is a fact of common experience that compact writing with the lines well spaced by elegant ascenders and descenders is more legible and economical than lines of fat sprawling letters separated by much smaller spaces.

Spacing of letters and words

Fanti's instructions are as follows: 'Now we have to discuss the distance between one letter and another. This should be equal to the width of the black parts of the 1etter. For example, the space between two n's should be the same as the width of its two legs when one is placed next to the other. Similarly the distance between words should be such that an o or n could fall between them.’ Tagliente suggests a somewhat wider space between words, advising his readers to ‘see that the distance between letters always equals the space between the legs of n...the distance between one word and the next should equal the space occupied by the letter m’. Vincento and Palitino have a similar rule, the former writing: ‘Make sure ,.,that the distance between one word and another is equivalent to the width of an n and that, as you join one letter to another, the distance between them equals the width of the white space between the legs of n’. So, too Augustino: ‘The space which should separate one letter from another should be equivalent to the white space between the legs of a well-shape n. The interval or spacing between one word and the full width of the latter u. These rules should be applied according to the judgement of the eye and not always by actual measurement because it would be tiresome to measure every letter as you write.' Ytiar also relates his spacing to the letter n but to an extraordinarily narrow version of the letter. He writes: 'The space between each letter should equal the distance between the legs of the n. And if someone should object that this still does not determine it, because the space in the middle of an n is not certain as have not so far mentioned the point, I reply that the correct proportion of the white space between the legs of the n is that it should be as wide as the thickness of one of its legs. . . . The distance between words should equal twice the white space of an n, that is to say, that the space between words is double the space between letters.' Mercator offers no guidance about word spacing, but his rules for letter spacing are more sophisticated – ‘the distance between individual letters should be the same as that between the two downstrokes of y, but the distance between c, e,f,g, r, t, and v and the succeeding letters should be half of this.'

Capital letters

In forming the capital letters the discipline imposed by the canons which regulated the writing of minuscules was to some extent relaxed in favour of personal expression. Writing masters usually took this opportunity to display their individual skill, elan, and inventions. Thus we find a pleasing contrast between the legible regularity of the text and the freedom of the capital letters. The price was that writing masters could not, perhaps would not, give detailed rules for the construction of majuscules. Thus Fanti merely refers the student to his models, which, as already mentioned, were never printed, while Tagliente keeps off the subject. Vicentino says: 'Write your majuscules so that they always stand upright, making then1 with confident, well-defined strokes with no shakiness about them. . . It should not be a great effort for you to learn all the majuscules once you have trained your hand with the minuscules. . . I will say no more than that you should try to learn to shape the majuscules as you find here in the example which I have written for you.' Nor is Palatino more explicit, merely stating: 'The chancery majuscules are all derived from the same three basic strokes as the minuscules. But because there are no fixed rules for them, they are made according to the judgment of the eye. Note that the strokes should be lively and confident, with no shakiness. Augustino says that the capital letters 'do not have any rigid rules governing their construction because they are written in various sizes and fashions according to the way which the quill is cut. They can be formed with a variety of strokes and they are made more rapidly with the eye than by any other method. They should always have lively confident strokes, not spoiled by shakiness or any other fault, if you want to construct them properly. . . copy the models that you see here'. Yciar, who relied extensively on earlier writers for much of his work, is at a loss: 'I declare that none of the authors who have come to my notice has so far given rules for the construction of these majuscules. Battista Palatino, the most recent writer, says that there is really no fixed rule other than that you should make them according to the judgment of your eye, copying the model alphabet which he has provided and taking pains to fashion the strokes lightly with a very steady hand.' Mercator is a trifle more helpful: ‘Capital letters correspond in dimension with their respective minuscules. They have the same slope to the right but can be perpendicular if preferred. Sufficient examples will be found throughout this manual and in the alphabets that follow.'


The question of joins is closely related to the proportions of letters, their spacing, and the angle at which the pen is held. Let us assume that the width of, say, letter n is four units (such as pen-widths), the legs accounting for two units and the white space between for the other two units. These are typical proportions for the classic chancery hand. Let us then apply the writing-masters' canon for letter spacing, ie, the width of the white space between the legs of n (or two units). If we construct on the base of four units (1) letters n in the proportion 1:1, i.e. four units high; (2) letters n in the proportion 2:3 (Mercator's), i.e., six units high (3) letters n in the proportion 1:2 (Palatino’s and Augustino’s) and line up each series of n's so that a space of two units separates one from the other, it becomes immediately clear that the type of join is different in each example. In case (1) we can join the two n's diagonally without too much distortion, but the space between letters is too generous. Case (2) demands a sharper angle and therefore a thicker stroke. In case (3) the angle is so steep that the writer will almost certainly twist his pen in writing and this stroke itself will become so wide as to impair the contrast between thick and thin which is so characteristic of the italic hand. There will also be a tendency to try to make room for the join by over-sloping the same letter. It is surprising how often even the best scribes do this. The word in provides a good test. The choices boil down to: (a) letter spacing that will lead to a sprawling script (popular with composers of contemporary manuals, with a few honourable exceptions), (b) attempting to join at the cost of spoiling the rhythm of writing and sometimes the shape of the succeeding letter, (c) preserving moderate spacing and letting the letters join as they wish; in a natural fashion - a kind of free love, if you like. On the whole the Italian writing masters tended to adopt this third solution, though not entirely.

Fanti offers little guidance. He says that the joins are made with the flick which ends such letters as a. When constructing it, you must give the letter a little of a certain movement which is called the' dead line' (linea marta) because it is almost invisible. Elsewhere he compares it to 'the tail of a lion'. Note that the stroke is a short one: it cannot be used when the letters are either too far apart or too tall. Tagliente describes the stroke as a 'graceful little turn which is called the "finishing stroke" because it finishes off the main stroke.' He puts forward his scheme rather inadequately by illustrating how the word magnifico is written: 'First make an m with its finishing stroke. Lift your pen and make an a next to the m (thus giving us "ma"). Starting from the finishing stroke of a, make the upper bowl of the letter g. After you have completed g, pick up the opening stroke of n from the bowl of g, and make the letter n. Without lifting the pen from the paper, continue the finishing stroke of n and make the letter i (this gives us 'magni"). Next write the letter f and next to it the letter i in one movement (this' gives us "magnifi"'). Similarly continue the finishing stroke of i and write the first circular element of c, then add to c its head piece.[4] Next make the letter: o beside c (and this gives us "magnifico").' The student is advised to go on Joining and linking all the letters above and below as far as you can without lifting your hand, unless you have to, until you have completed the word.' Although Tagliente's language is imprecise ,he seems to be saying that it is not necessary to try to join each letter but to allow the letters to meet as convenient.

We get more detailed instruction from Vicentino, whose proportion for small letters is roughly 3:5. He writes: 'As regards the small letters of the alphabet, some can be joined to an immediately following letter and others cannot. Those that can be joined to a succeeding letter are the following: a, c, d, f, i, k, 1, m, n, f, s, t, u. Of these, a, d, i, k, 1, m, n, u may be joined to every succeeding letter: but c, f, f, s, t can only be joined with some. The remaining letters of the alphabet, i.e., b, e, g, h, o, p, q, r, x, y, z,[5] must never be joined to the letter that follows. Nevertheless I leave it to you to decide whether to join or not, provided you keep your letters even.' The last six words are the acid test.

Vicentino then gives an example showing a joined to every other letter of the alphabet. In some cases, the join cannot be made without being lengthened in inadequate space and distortion follows. The junction of im is a case in point. Thus Vicentino, by his own example, makes it doubtful whether the letters that end like a should be joined to every succeeding letter. Vicentino concludes with advice for joining d, f, s, f, and t.

With his 1:2 proportion for letters, Palatino has to face the problem in its acutest form. He commits himself to finishing letters like a 'with a short upward diagonal stroke. This stroke serves to join and link one letter with another'. He then states: 'As regards the joining of one letter to another - since other authorities have written about it at great length and really in a very confused manner - I will give you this concise rule: all the letters which end in a little diagonal or a flick of the pen, such as a, c, d, e are joined to immediately succeeding letters. . . f and t are joined [by the crossbar] to all letters without ascenders.' Palatino preserves the canon of spacing and the angle of his pen. In practice, therefore, most of his joins touch the succeeding -letter quite low down. In principle, he is keeping to Tagliente's rule of comfortable contact.

Augustino, whose letters are constructed to Palatino's 1:2 proportion, gives four rules: 'All letters which begin with a diagonal serif can, and should, be joined with the preceding and succeeding letters. . . but those that end in a crossbar, such as f and t, are joined by the crossbar to the succeeding letters. The following letters, b, g, h, o, p, q do not themselves make a join, but they may be joined by other letters. The following characters, x, y, z cannot be correctly attached to other letters.'

Yciar refers to the 'little serif', which completes letters like a, and 'serves to connect one letter with another'. Then, in somewhat fanciful language, derived apparently from Erasmus, he writes: 'Some letters are on such bad terms with each other that they absolutely refuse to join in any friendship or intercourse with others: for example, g, h, o, p, e. . . . Other letters are affectionate and sociable by nature and, so far as they can, they do not withhold their intimacy from any other letter. Such are all those that end with a diagonal serif, c, d, e, 1, m, n, t, u, etc. . . . The second rule is that f and t can be joined by the cross-stroke to succeeding letters.’

Mercator teaches that 'all letters that end in a serif rising from the lower line of writing should be joined, by prolonging the serif, to those that follow, with the exception of c. . . . So e, c, and t, although they finish with a hairline at the foot, cannot be joined to the following letter. So long as a crossbar does not follow, letters are connected with a single stroke of the pen. . . but when a crossbar follows, the diagonal is prolonged from the previous letter to a point a little beyond the centre of the lines of writing: then you must transfer the pen to the upper line'. Because of the 2:3 proportion of his small letters, Mercator has relatively more room for his diagonal joins. He prolongs them, but not as far as the top line of writing; even this is sufficient to give his models a strong diagonal movement.

The real proof of a system of joins is whether it allows the even spacing of letters standing compactly together. Unless the emphasis is put on preserving sound letterforms and this sort of spacing, joins will tend to be excessive and are likely to result in writing that is either squashed up or (more probably) flaccid, floppy, and sprawling. It slows up the writing and makes it more difficult to read quickly. Excessive joining can also become a mama. In one recent model, for example, which is claimed to be 'italic,' the descender of q has been truncated, simply to facilitate a pointless join with the succeeding u.


Such were the canons that regulated the classic chancery hand. In our own day, this hand has been revived, under the name of 'italic', as a successful solution to the problems of modern handwriting. The italic models which approach most nearly to the spirit and effectiveness of the original are those of Alfred Fairbank as seen in his Handwriting Manual and Beacon Writing Series. Many of those who have subsequently tried to attract attention in this field have found it convenient to describe their wares as italic when they are manifestly not consistent with the primary canons. This is not to say that such models are bad: only that they are sailing under false colours. The commonest deviations are sprawling over-wide letters (leading to over-wide spacing), excessive slope, over-stressed, redundant joins which break the rhythm of writing and impair spacing, and miscellaneous fads, presumably intended to demonstrate the author's originality.

[1] Authors still write with monotonous regularity that this was 'the first Italian writing-book' and that it came out in 1522.

[2] Shorter ascenders and descenders might be used for copying manuscripts in a bookhand.

[3] In II Secretario (Venice, 1581), Marcello Scalzini wrote: 'They can be readily reduced to suit a smaller letter, whereas short ascenders and descenders, once learnt, are difficult to reduce or lengthen gracefully'.

[4] Here c is made in two strokes like e, so neither letter is suitable for joining to the next.

[5] h, when made by curving its second .leg in a little at the bottom, does not have the diagonal finishing stroke and is, therefore, not joined.