Line, Shape, Texture, Pattern

Your brain sees trees, walls and people. The camera sees a bright area, a dark area and a diagonal line. Those two ways of looking produce completely different photographs.

What this chapter trains is the ability to switch object recognition off for a moment and look graphically instead. The question to ask on location is:

If I take away the colours and the names, what is left in this frame?

Whatever remains — a few lines, some shapes, a patch of texture, a run of repetition — is what the camera is actually going to record. It is difficult at first, and once the switch happens a surprising number of compositional problems solve themselves.

Seeing graphically is not an artistic accomplishment

A misconception worth dismantling first: many people assume that seeing things as lines and shapes is a talent, or a layer of taste laid over ordinary perception.

It is the reverse. Seeing graphically is reverting, not refining.

The first thing your visual system does when identifying anything is find edges. Cells at the front of the visual cortex respond to light–dark boundaries at particular orientations — some only to vertical edges, some only to forty-five degrees — so the whole sheet amounts to a bank of orientation detectors. “A tree” is a conclusion that surfaces several processing stages later, and the raw material at the bottom of that stack is the same thing the camera records: a distribution of light and dark.

So when you say you see a tree, you are reporting a heavily processed conclusion. The camera has none of those stages. It stops at the first one. Seeing graphically means stepping back to where the camera already is.

The visual system does something else that makes lines more powerful than you would guess: it actively exaggerates edges. Neighbouring cells inhibit one another, so at a boundary the bright side is pushed brighter and the dark side darker. You can verify this yourself with a row of grey patches running light to dark: each patch is uniform, yet every one appears to be a different tone at its left and right edges. Those are Mach bands, added by your visual system and absent from the world.

The practical consequence: any clear tonal boundary in a frame gets amplified before the viewer is aware of it. The force of a line is not a matter of opinion — it has a physiological source. It is also why a cleanly composed photograph feels strong. Not because of the subject, but because a handful of amplified edges have not been drowned in noise.

Line

Lines have direction and direction carries feeling. That is not mysticism; it is that each direction corresponds to something your body knows.

Direction	Bodily experience	Reads as
Horizontal	Horizon, lying down, rest	Settled, slow, expansive
Vertical	Standing, trees, buildings, gravity	Strength, formality, gravitas
Diagonal	Falling, accelerating, a slope	Unstable, dynamic, tense
Curved	Rivers, bodies, wind	Soft, unhurried, guided

Diagonals deserve separate treatment, because they are the most effective and the most easily overlooked. Horizontal and vertical are this world’s defaults — buildings, horizons, trunks, tabletops, all of them. A frame with no diagonal in it will be calm, and may be inert. Introduce one and the whole picture wakes up, precisely because it is the one thing contradicting the environment’s default directions.

The fastest way to get one is not to hunt for it but to manufacture it: crouch so the ground line tilts, move to the side of a staircase or ramp, wait until a low sun lays a shadow across at an angle, or simply rotate the camera thirty degrees. The first three are diagonals of content. The last is cheap and wears out quickly, but it does prove that direction alone changes how a frame feels.

Note too that which way a diagonal runs matters. For anyone who reads left to right, a line rising from bottom-left to top-right reads as ascent; the mirror image reads as descent. There is no need to be superstitious about it, but it explains why the same hill feels different photographed from either side.

Shape and silhouette

There are two kinds of shape, and the light decides which one you get — not the object.

Flat shape is an object’s outline: a two-dimensional block of black or colour with no internal information. Volume is that same object rendered as something thick, weighty, occupying space, and it is carried entirely by the gradient from light to dark across its surface.

Which one a sphere becomes depends on where the light is. Frontal light presses it into a flat disc — shape. Side light lays a gradient across it — volume. Backlight turns it into a pure black circle — shape at its most extreme. So “this photograph looks flat” and “this photograph looks solid” is not a difference of technique. It is which light you chose.

That yields a direct decision on location: settle first on whether you want outline or volume. For outline, find backlight and put all your attention on the contour. For volume, find side light and watch that the gradient between light and dark is not being crushed. A frame that wants both usually gets neither.

When there is not enough light to render detail, an object collapses into shape. This is one of the most valuable uses of backlight: it simplifies something complicated down to its contour.

Two things decide whether a silhouette works:

• The outline has to be identifiable. A silhouette contains no detail whatsoever, so the contour is the only information the viewer gets. Choose the most characteristic angle — profile beats frontal almost every time. A person’s frontal silhouette is a post; the profile has a nose, a chin, a haircut.
• Outlines must not overlap. Two silhouettes touching become one black mass, because there is no tonal difference available to separate them. That is what happened at the left of the frame above.

On location the second has exactly one fix: move until every subject’s outline stands clear. Usually two paces will do it. But you have to know what you are looking at first — if you see “a group of gazelles” you will not notice that two have merged. Only when you see “a set of black shapes” does the problem announce itself.

Silhouettes are not all-or-nothing, incidentally. Retaining a trace of shadow detail — contour clean, a little internal structure still visible — usually holds attention longer than a dead-black cut-out. The price is that exposure has to be more precise.

Texture

Texture depends entirely on the angle of the light. One sentence covers it: texture is made of shadows, not of resolution. Forty megapixels will not render a rough wall under frontal light. Twenty will, under low side light.

You can calculate this, and having calculated it you know where to stand. Take a ridge two millimetres high on a wall:

• Light grazing at thirty degrees casts a shadow about 2 ÷ tan(30°) ≈ 3.5 mm. Visible, not emphatic.
• Light grazing at ten degrees casts about 2 ÷ tan(10°) ≈ 11 mm — more than five times the height of the ridge itself, and the texture is unmistakable.
• Light from the front, near ninety degrees, casts a shadow tending to zero. Texture gone.

So what texture actually requires is not hard light but a grazing angle. A soft light arriving very low — early window light from the side — will out-render a hard light coming straight down. This is the same fact as the lighting chapters, seen from the other side.

Which brings out a second practical point: one texture alone gets dull quickly; two together have content. Coarse against fine, regular against disorderly, hard against soft — contrast is what makes each of them legible. An evenly rough wall filling the frame reads as a material sample.

There is a trick worth stealing from painters here. Squint until the scene is barely in focus. Fine detail disappears and only the large masses of light and dark survive — which is very close to what the photograph will look like from across a room. If a frame falls apart when you squint at it, no amount of sharpness will save it.

Pattern, and the break in the pattern

Repetition produces pattern and pattern produces rhythm. But perfectly even repetition palls fast — the brain solves it within seconds and then has nothing left to do. This is the fixation argument from the previous chapter again: once the rule is understood, none of the remaining units needs looking at, and the eye is out of work.

So there are two routes to photographing pattern, and they are genuinely different things.

The first is regular irregularity: repeating in structure, every unit different in detail. Log stacks, roof tiles, cobbles, crowds. It holds attention because the brain never finishes solving it — the regularity invites you to skim, the detail keeps stopping you.

The second is regularity plus one exception: keep the repetition as even as you can and put a break inside it. One tile of a different colour, one open window, one person turning their head, one stack of loungers in the wrong place. The exception becomes the focal point instantly, and the regularity all around pushes it forward.

The strength of this scales with two things: the more even the repetition, the more the exception stands out, and the smaller the exception, the more concentrated the effect. So composition here means doing two opposite things at once — excluding anything that spoils the regularity, while making sure the exception itself is not clipped by an edge.

There is a third use, which the frame above demonstrates: let the repetition run past every edge. Once no border is visible the viewer loses the cues for scale, and a three-dimensional building is flattened into a two-dimensional plane. It is the extreme version of this chapter’s opening sentence — you have successfully turned a block of flats into a set of rectangles.

Six symptoms and their causes

The format throughout: the picture shows X → the usual cause is Y → change Z.

1. Walls, wood grain and old hands come out looking painted over. → Light too frontal; no small ridge has a shadow of its own. → Move to where the light grazes the surface; remember ten degrees beats thirty by a factor of five.
2. A silhouette is an unreadable black blob. → Two or more outlines overlapping, or a frontal angle chosen. → Move until the outlines separate, and shoot the profile.
3. A pattern picture you do not want to look at twice. → Perfectly even repetition, solved in seconds. → Put an exception in it, or photograph the kind of repetition whose detail never repeats.
4. There is an exception in the frame and nobody notices it. → It is clipped by an edge, or the surroundings are not regular enough. → Move it inward and take the clutter out of the frame.
5. The picture is calm and you wanted it lively. → Nothing but horizontals and verticals; no diagonal anywhere. → Crouch, change angle, or find a low-angle shadow.
6. The texture is there but the frame looks messy. → A single texture filling the picture with nothing to compare it against. → Introduce a second texture, or keep a smooth area as a foil.

Somebody taught this

Seeing the world as line and shape sounds like personal style. It is in fact a teachable tradition with names attached.

Paul Strand got there first, in 1916. Photographing bowls on a table and a porch railing from directly above, he pushed everyday objects right to the edge of unrecognisability; the pictures ran in the final issue of Camera Work. What he demonstrated is the whole premise of this chapter — that a photograph can be organised by its shapes rather than by its subject, and still be a photograph rather than a diagram.

László Moholy-Nagy turned it into a curriculum. At the Bauhaus in the 1920s he argued that photography was not for reproducing what the eye sees but was a distinct “new vision”, and set students to shoot steeply down, steeply up, and to make photograms with no camera at all — every exercise designed to sever the act of recognising the object. The training’s core is this chapter’s opening question: read the distribution of light and dark first, and only then ask what it is.

Karl Blossfeldt did the other half. He set out to prepare teaching material on plant form for a design course, photographing ferns, seed pods and tendrils enlarged many times over — and the results look like wrought-iron railings, capitals, architectural components. What he proved is simple: magnify anything past the point of recognition and it reverts to shape.

Edward Weston photographed peppers until they stopped being peppers. His 1930 close-ups are still, unmistakably, a vegetable on a sheet of paper, and simultaneously a study of volume — the whole picture is the gradient wrapping around a form. The pepper is the argument in this chapter’s shape-and-volume section, made forty years before anyone wrote it down as a rule.

A field exercise table

What you see	As graphics it is	What to do
A row of trees, railings, columns	A set of parallel verticals	Find a diagonal or a curve to break them
A backlit person or animal	A set of black shapes	Move until the outlines separate; shoot the profile
A rough surface	Small-scale light and dark	Wait for the light to graze it, or move to the side it grazes
Things in neat ranks	Pattern	Find the exception; if there is none, find repetition whose detail varies
Clutter you cannot photograph	Too many lines, no dominant one	Move in until one or two elements remain, or raise/lower the viewpoint to cut layers
“This felt great and photographed badly”	Colour or atmosphere was doing the work	Switch to a monochrome preview and look again — what survives is what was real

That last row is the most useful trick in the chapter. Both the X-T5 and the X-E4 will set the film simulation to ACROS or MONOCHROME and show a black-and-white image in the viewfinder. With colour removed, line, shape and tonal distribution surface immediately — which is precisely what this chapter has been asking you to look at. Shoot raw and none of the colour information in the file is lost.