Pre-Writing Strokes Before Letters

Children copy shapes in a stable order: vertical line, horizontal line, circle, cross, square, single diagonal, X, triangle. The cross is the useful checkpoint. It is the first form that requires a child to finish one stroke, lift, and place a second stroke in a planned relationship to the first, which is exactly what every letter demands and what no single line or circle has asked of them yet.

If a four year old can draw a circle and a line but cannot copy a plus sign from a printed model, letter worksheets will not teach them letters. They will teach them workarounds.

The order the shapes arrive in

The sequence is not arbitrary and it is not something you can reorder by drilling. Each form adds one new motor demand on top of the last. Standardized copy tests are built on this: the Beery-Buktenica Developmental Test of Visual-Motor Integration presents forms “arranged in order of increasing difficulty”, starting with lines and ending with forms most children never draw spontaneously.

Ages below are rough. Some are sourced, some are practitioner convention, and I have marked which is which rather than presenting all eight as if they carried equal weight.

Shape	Typically copied around	What it unlocks in letters	How well sourced
Vertical line	2 to 3	l, i, t, and the down stem of b d h k m n p r u	Convention. No public norm table found.
Horizontal line	2 to 3	the crossbars of t and f, the arms of E F H L	Convention. No public norm table found.
Circle	3	o c a d g q e	MedlinePlus lists “can copy a circle” at age 3
Cross (a plus sign)	around 4	the intersections in t f x, and in E H	Convention. Sits between circle and square in standard copy-form order.
Square	4	corners, closed shapes, the frames of E F H L	MedlinePlus lists drawing a square at age 4
Single diagonal	4 to 5	v w y z k, and the legs of A M N	Convention.
X (two crossed diagonals)	4 to 5	x, and the crossing in k and y	Convention.
Triangle	5	A K M V W	MedlinePlus lists “can copy a triangle” at 5; the AAP lists “copies triangle and other geometric patterns” in the 4 to 5 band

What actually changes at each step

Vertical and horizontal lines are one motor plan each. The hand travels in a single direction and stops. Nothing has to be aimed.

A circle adds continuous direction change and a return to the starting point. The hand never stops but never travels straight either.

A cross adds the thing letters are made of: two separate strokes, where the second one has to be aimed at the first. The child has to hold the position of stroke one in mind while executing stroke two. That is a planning task, not a muscle task, and it is why the cross is a better readiness signal than a neat circle.

A square adds three corners, each a full stop and a ninety degree turn, plus a closure back to the start.

Diagonals are the last to come because they have no anchor. Vertical and horizontal have gravity and the paper edge to reference. A diagonal has neither, so its angle has to be generated internally and held. The common explanation is that diagonals require crossing the body midline, which is plausible but not something I would state as established mechanism.

The triangle is last because it is three diagonals, three corners, and a closure all at once.

Why the cross matters more than the circle

Letters are not shapes. They are sequences of strokes assembled in a specific order with specific relative placement. An Australian pilot trial of a prewriting program put it directly: prewriting patterns are “the pencil strokes a child learns that are later used to form letters”, and “as letters are formed using a combination of prewriting patterns, the automatic recall of prewriting patterns may be required prior to the integration of these patterns into letter formations” (Johnston et al., 2025, Australian Occupational Therapy Journal).

A circle proves the hand can move. A cross proves the child can sequence and aim. Handwriting needs the second one.

When you skip that check and hand a child letter sheets anyway, they still produce something, because children are resourceful. What they produce is usually a set of compensations: letters drawn bottom to top, strokes retraced two or three times to reach the line, capitals substituted for lowercase because capitals are mostly straight lines, and letters assembled in whatever order gets the shape closed. Those habits are durable and slow to unpick later. Some b and d confusion at this stage is developmentally ordinary, but a child who has invented their own stroke order for twelve letters is a harder problem than a child who simply has not started yet.

What to do about it

Test before you teach

Copying is not the same as imitating. If you draw a cross while the child watches and they then draw one, that is imitation, and it is easier. A copy task means you show a finished cross already on the paper, say nothing about how it was made, and ask them to make one like it. Use the copy version. It is the harder and more informative test.

Run through all eight forms in about two minutes, one per sheet of paper, largest to smallest. Note the last one they get cleanly. That is your starting point.

Build the missing strokes without a pencil

At this stage the surface matters more than the tool, and a pencil is often the worst available option. If grip is still unsettled, forcing pencil work now just adds a second problem.

• Vertical and horizontal: work on a vertical surface. Paper taped to a wall, an easel, chalk on a fence, a wet paintbrush on dry brick. Vertical surfaces put the wrist in extension and take the shoulder out of the equation as a source of error.
• Circles: anything continuous and unbroken. Stirring, winding, spirals drawn as big as the arm reaches, snail shells, spinning a lid.
• Cross: give the second stroke a target. Roads that meet, railway tracks, tic tac toe grids, an X marking treasure on a map, a plus sign built from two lolly sticks before it is drawn.
• Square: boxes, windows, gates, frames drawn around a finished picture.
• Diagonals and X: mountains, rooftops, zigzag paths, the flight line of a thrown ball, a ladder leaning on a wall.
• Triangle: leave it. It arrives on its own once diagonals and corners are solid.

Keep sessions to about five minutes and keep them big. Big movements are easier to plan and easier for you to see going wrong. Shrinking to worksheet size is the last step, not the first.

Cross into letters with the letters they can already make

Once circle and cross are reliable, do not open at A. Start with the letters that use only strokes the child already owns: l, i, t, o, c. That is one vertical, one horizontal, and one circle, recombined. There is a fuller case for this in why the alphabet is the wrong teaching order, and the full letter index is grouped so you can pick by stroke rather than by alphabet.

Set the worksheet generator to the largest line size and dotted tracing, put four or five letters on a page, and stop there. Diagonal letters (v, w, x, y, z, k) wait until diagonals are copied cleanly, which for many children is a year later.

Where the evidence is genuinely thin

Three honest caveats, because this area attracts more confidence than it has earned.

First, shape copying predicts later school performance but is not the same thing as handwriting. Design copy skill measured before kindergarten predicted higher achievement at kindergarten entry and greater gains across the year (Cameron et al., 2012, Child Development, doi:10.1111/j.1467-8624.2012.01768.x), and fine motor integration in preschoolers, measured partly by copying shapes and letters, predicted later literacy and mathematics competency (Escolano-Pérez et al., 2020, Frontiers in Psychology, doi:10.3389/fpsyg.2020.01044). Both are correlational. Neither shows that drilling shapes causes better handwriting.

Second, the most common copy-form test is not a handwriting test. Reviewing the Beery VMI as an outcome measure, Pfeiffer and colleagues found that “the VMI measures a related but different construct from handwriting” and that it “was never intended to assess handwriting ability, nor was it designed to screen specifically for handwriting dysfunction” (2015, American Journal of Occupational Therapy, doi:10.5014/ajot.2015.015826).

Third, direct prewriting instruction has not clearly beaten ordinary classroom practice. The 2025 pilot trial above ran six weekly sessions with eighteen four and five year olds. Everyone improved over time, but the intervention group did not improve significantly more than the waitlist control, and the authors flagged small sample size and possibly insufficient dosage as reasons.

The practical reading: use the stroke sequence as a readiness check and as a way of choosing which letters to start with. Do not treat it as a curriculum that has to be completed before a pencil is allowed. If a child is five and a half with no reliable circle, or shows no interest in making marks of any kind by four, that is worth an OT evaluation rather than more practice sheets. For the wider picture of what hands can do at each age, see fine motor milestones for handwriting, and for the sequence once letters begin, how to teach handwriting step by step.