Landscape design

Why a garden with good plants can still look wrong

Most gardens that disappoint their owners are not short of healthy plants. They are short of relationships between them. Every plant is fine on its own and the whole reads as a collection rather than a place.

Design is the set of relationships: how sizes relate, how shapes repeat, where the eye goes and where it rests. None of it requires drawing skill, and none of it is about taste. It is closer to a small number of rules that hold whether the garden is formal, wild, tiny or two acres.

This page is the principles. If what you need is the practical sequence of building a garden, starting from drainage and levels, that is a different question and it is covered separately in the backyard planning entry.

Balance and unity, the two principles that are hardest to name

The six principles above are the ones that translate directly into an action. Two more sit underneath them, harder to point at and easier to feel, and every published account of landscape design lists them.

Balance

Balance is visual weight distributed either side of a view. It comes in two forms and the choice is a commitment.

• Symmetrical balance mirrors one side against the other: matched beds, matched paths, matched plants. It reads as formal, it is immediately legible, and it costs more in maintenance, because the effect fails the moment one of a matched pair dies or grows differently from its twin.
• Asymmetrical balance counterweights instead of mirroring. One large mass on one side against several smaller ones on the other, arriving at similar visual weight by different means. It is more forgiving, more relaxed, and considerably harder to get right the first time.

The common failure is neither: a border that started symmetrical and drifted, so one side is formal and the other is a collection. Pick one and commit, and if in doubt look at the view from the window you use most rather than from the middle of the lawn.

Unity

Unity is what makes several parts read as one place instead of as several unrelated gardens joined at the edges. It is the principle repetition serves, and it works on more than plants.

• Repeat a material. The same paving or the same edging appearing in two separate areas ties them together more strongly than any planting will.
• Repeat a color or a shape through areas that are otherwise different in character.
• Transition gradually. Where two areas differ in scale, step between them: small trees to shrubs to herbaceous plants to groundcover, or coarse texture to medium to fine. An abrupt jump reads as a boundary whether or not you built one.

Unity is also the reason a garden of individually excellent plants can look wrong, which is where this page started. Every plant is fine; nothing repeats; and the eye finds no thread to follow.

Diagnosing a garden that looks wrong

The quickest route into the principles is backwards, from the symptom. Stand where you usually stand and name what bothers you, then read across.

Almost every entry in that right-hand column is subtraction or repetition rather than purchase, which is the most useful thing to know about garden design and the least often said.

The front yard, where most of this is decided

For most American properties the front yard is the whole of what anyone else sees of the landscape, and it is also where the scale principle is broken most reliably. It is worth treating as its own problem.

Its job, from the public-area definition above, is to make the way in obvious and to present the house. That is a short list of requirements and it is why front yards look best with less in them rather than more.

Foundation planting, and the overplanting trap

Clemson names overplanting as probably the most common failure of foundation plantings, and it happens for an understandable reason: a bed of small shrubs at their mature spacing looks sparse on the day it is planted, so people fill it. Five years later the same bed is a solid mass covering the windows.

• Leave an air gap at the wall. Plants should reach their mature width and still leave 12 to 18 inches of air next to the wall, and shorter shrubs want at least a 3-foot clearance from the house for air circulation and foundation vents. Nothing goes within about a foot and a half of the foundation anyway, because the soil there is too dry.
• Space to 70 percent. Set shrubs so they begin to merge at roughly 70 percent of mature size. That is the spacing that looks thin at planting and right for the following two decades.
• For spreading shrubs, allow a width equal to the eventual height. A useful shortcut when a label gives one figure and not the other.
• Never put a tall plant under a window. It will cover it, and then it will be sheared into a box every year to stop it.
• Keep young trees away from the house. Planted too close, a tree never develops a proper shape, and it is the shape you were buying.

The consequences of getting this wrong are not only visual. Overplanted foundations cut the airflow along the wall, which brings algae on siding, and branches rubbing the house scratch it. Aggressive roots hard against a footing are a genuine risk on the wrong species.

What a front yard actually needs

Working from the principles rather than from a catalog, the list is short.

• A route to the door that reads as one, wide enough for two people, going where people would walk anyway.
• One focal point, which in most front yards is the front door itself. Everything else supports it rather than competing.
• Repetition along the frontage, since this is the view most often seen from a distance and at speed, where repetition reads and variety does not.
• Structure that works in winter, because a front yard is looked at every day of the year rather than sat in on summer evenings.
• A tree in scale with the house, sited far enough out to grow into its shape.

Almost nothing on that list is a flower, which is the honest summary of front-yard landscape design and the reason so many of them disappoint the people who spend the most on them.

Reading the site before designing it

The principles are about relationships between things. What decides which things are possible is the site, and extension guidance is consistent that this comes first: most site characteristics are difficult and expensive to alter, so the sensible order is to record them and design with them.

A base plan is a scaled drawing of what exists: boundaries, the house and its doors and windows, drives and paths, utilities, and every tree and shrub staying. Onto that go the observations.

• Sun, by hours. Mark how many hours of direct sun each area gets, because that is the number plant labels are written against. Full sun means 6 to 8 hours or more, partial sun or shade 4 to 6, and shade under 4. A single afternoon of noting this settles arguments that otherwise run for years.
• Wind, and where it funnels. Between buildings and along fence lines is where it accelerates.
• Water. Where it stands after rain, where it runs, and where downspouts discharge.
• Soil, tested rather than assumed.
• Views out, and views in. What you want to see and what you want screened, marked as arrows from where you actually sit or stand.
• Sounds and nuisances. Road noise and a neighbor’s security light are design problems the same as a bad view.

The step people skip is the one between the survey and the planting: decide what each area is for before deciding what goes in it. Extension publications divide a property into three use areas, and the division is more useful than it sounds.

• Public, usually the front: the approach, the entrance, where visitors arrive and park. Its job is to make the way in obvious.
• Private or family, usually the back: sitting, eating, play, the garden you use.
• Service: bins, compost, storage, the heat pump, the washing line. Every property has one whether or not it was planned, and a garden that never allocated it ends up with these things in the middle of the private area.

Rough bubbles on tracing paper over the base plan are enough at this stage: circles for each use, sized roughly right, moved around until the traffic between them makes sense. Nothing is drawn to scale and no plants are chosen. Getting this right is what makes the planting decisions easy later, and getting it wrong is why some gardens never feel comfortable no matter what is planted in them.

Function before appearance is the summary. A well-designed landscape is both, but the pleasing part is much easier to achieve on a layout that already works than to retrofit onto one that does not.

Scale, and the mistake everyone makes

Scale is the relationship between the size of things and the size of the space and the people in it, and it is the principle most often broken.

The characteristic error is buying to the size of the pot. A three-foot shrub in a nursery container looks generous beside a bare bed and ridiculous beside a two-story house. Meanwhile a bed one foot deep against a fence cannot hold anything in proportion to the fence.

The correction is uncomfortable and reliable: make the beds deeper than feels right, use fewer and larger plants rather than many small ones, and check every purchase against its mature dimensions rather than its current ones.

Repetition is what makes it look designed

If there is one principle that separates gardens that look considered from gardens that look accumulated, it is repetition.

Three of the same shrub spaced along a border reads as a decision. Nine different shrubs in the same run reads as a plant collection. The eye finds the repeated element, follows it, and reads the whole planting as one thing.

This is also the cheapest principle to apply, because it costs nothing extra and usually less: three of one plant is a smaller bill than nine of nine.

Repetition works on anything, not just plants. The same paving material used in two places, the same pot shape, the same color appearing three times down a border. It is what unifies a space without making it uniform.

Structure before decoration

Every garden has a permanent layer and a changing one. The permanent layer is the trees, shrubs, hedges, walls, paths and levels. The changing layer is the perennials, bulbs and annuals.

Most amateur gardens are heavy on the changing layer and light on the permanent one, which is why they look wonderful in June and empty in January. A garden should be legible with every herbaceous plant cut to the ground.

The test is simple: look at a photograph of the garden in February. If there is nothing there, the structure is missing, and no amount of summer planting substitutes for it.

Give the eye somewhere to go

A space with nothing to look at is uncomfortable, and a space where everything competes is exhausting.

One focal point per view is the working rule. A tree, a seat, a pot, an opening in a hedge, placed where the eye naturally travels: at the end of a path, at the turn of a bed, in the line of sight from the kitchen window.

The corollary matters as much. Everything else in that view supports the focal point rather than competing with it, which means resisting the urge to put something interesting in every corner.

Hide the whole from the beginning

A garden you can see all of at once is a garden you have finished looking at within seconds, and this is why small gardens often feel smaller than they are.

Interrupting the view makes a space feel larger, and the technique is old and reliable: a screen, a large shrub, a change of level or a narrowing two thirds of the way down, so that part of the garden is implied rather than shown. The visitor knows there is more, and the space reads as bigger than its measurements.

Line, and where people actually walk

The lines of a garden are its edges: paths, bed shapes, the boundary between grass and planting. They are what the eye follows before it notices any plant.

Two rules cover most of it. Make the line deliberate, whether that is straight or curved, because a wobbling edge that is neither reads as an accident. And put paths where people already walk, because the worn line across the grass is the truth about the route, and a more elegant path laid elsewhere simply grows a second, informal one beside it.

Color comes last

Color is what most people start with and it is the least structural decision in the garden.

It is worth putting last for a practical reason: color is temporary. A flower color lasts two or three weeks a year, while form, texture and structure are there for twelve months. A garden designed around foliage shape and plant form works all year and can carry any color scheme; a garden designed around flower color works for a fortnight.

When color is decided, restraint reads better than variety. Two or three colors repeated through a border looks intentional. Every color available looks like a seed catalog.

Frequently asked questions

What is the most common landscaping mistake?

Buying plants at the size they are in the pot rather than the size on the label. Second is using one of everything instead of repeating a few things. Between them they account for most gardens that look crowded and unresolved after three years.

How do I make a small garden look bigger?

Stop it being visible all at once. A screen, a large shrub or a change of level two thirds of the way down leaves part of the space implied rather than shown. Repeating the same plants along a border also lengthens a view.

How many different plants should I use?

Fewer than instinct suggests, repeated more often. Three of one shrub along a border reads as intention; nine different ones in the same run reads as a collection. It is usually cheaper too.

Do I need a drawn plan?

Not necessarily, but you do need the decisions a plan forces: where the paths go, what the permanent structure is, what each view looks at. A hose laid on the ground to test a bed shape does most of the same work.

Where should I start?

With structure and lines rather than plants and color. Get the permanent layer and the shape of the space right and the planting is straightforward. Do it the other way round and the garden is rebuilt twice.