Circumstantial evidence and inferences

Such inferences should be logical inferences from proven facts, not mere speculation or guesswork. R v Puttick (1985) 1 CRNZ 644 (CA)

Introduction

In most criminal trials, the prosecution asks the fact finder to draw an inference from circumstantial evidence. The vocabulary is technical. The reasoning is plain: drawing a logical conclusion from established facts.

You do this every day, often without realising. You hear a kettle whistle and conclude the water has boiled. You smell smoke and conclude something is burning. You feel a cold draft and conclude a window is open. You arrive home, find the back door open and the dog gone, and conclude the dog has run off. None of these is a guess. Each is a logical conclusion drawn from established facts. The same reasoning runs through every criminal trial. To see how, picture a closed room.

The room

Imagine you are sitting in a warm, dry room. The door is closed. There are no windows. You cannot see outside.

A person walks in.

They are wet. Their hair is damp. Their clothes are darkened with water. A small puddle forms at their feet on the wooden floor.

What has happened outside?

The natural inference is that it is raining. That is the first conclusion most people would draw, and they would draw it quickly.

But pause. Is rain the only reasonable explanation for what you can see?

A child could have thrown a water balloon. A neighbour could have turned a hose on the person as a prank. They could have walked through a sprinkler. They could have spilled a bottle of water on themselves on the way in.

Each of those alternatives is, on its own, a reasonable explanation. None is far-fetched. The evidence is consistent with rain, but it is also consistent with several other things. The inference of rain is open, but it is not safe.

Now imagine the same scene.

The same person walks in, equally wet. This time they wear a heavy rain jacket beaded with water. They carry a closed umbrella, dripping steadily onto the floor. Their gumboots have tracked wet prints across the room behind them. The puddle at their feet is larger.

What has happened outside?

The inference of rain is now considerably stronger. A child with a water balloon does not explain the rain jacket. A garden hose does not explain the umbrella. A spilled bottle does not explain the wet gumboot prints leading from the door.

Each piece of evidence, on its own, might still be explained away. Together, they all point in the same direction. The alternatives begin to strain. They require coincidence on coincidence: the rain jacket worn for some unrelated reason, the umbrella carried for some unrelated reason, the gumboots wet from some unrelated source, all on the same person on the same afternoon. The alternatives become harder and harder to maintain. The inference of rain is open, and it is now safe.

Circumstantial evidence and inferences

This is how inferences operate in a criminal trial.

Most criminal cases turn at least in part on circumstantial evidence. Circumstantial evidence is evidence that does not directly establish a fact in issue but invites the fact finder to draw a conclusion about it. The prosecution is rarely able to put before a fact finder direct evidence of every fact it must prove. More often, the prosecution asks the fact finder to infer a fact, such as intention, knowledge, or identity, from other facts that are established in the evidence.

An inference is a conclusion drawn from facts that the fact finder accepts as reliably established: see R v Puttick (1985) 1 CRNZ 644 (CA). It is a logical deduction from established facts. It is not a guess.

Circumstantial evidence and inference are two parts of the same reasoning. The "circumstances" are the proved facts. The "inference" is the conclusion the fact finder is invited to draw from them. The Supreme Court restated the law on how juries are directed in these cases in Iongi v R [2025] NZSC 191: whatever the inferences, the question that matters is whether guilt is proved beyond reasonable doubt.

The familiar image is strands in a rope. The Court of Appeal in R v Gee [2001] 3 NZLR 729 (CA) described it as part of the normal direction to juries in circumstantial cases. No single strand may be strong enough on its own. But several strands twisted together can carry weight that none of them could carry alone. The prosecution does not need to prove each individual strand beyond reasonable doubt. What matters is the strength of the rope as a whole.

Applying this to the second scene, the strands are the rain jacket, the umbrella, the gumboots, and the water on the floor. Taken alone, each strand can be cut. A water balloon explains the wet jumper. A forgotten umbrella explains an umbrella in someone's hand. Wet gumboots can be explained by a leaking pipe. But four strands twisted together, all running in the same direction, are far harder to cut.

The fact finder looks at the evidence as a whole and asks whether they are sure of the defendant's guilt on the totality. Where proof of an essential issue depends on an inference from circumstantial evidence, the fact finder must be sure that the inference the prosecution advances is the only reasonably available one. A strong probability of guilt is not enough: R v Maxwell (1988) 3 CRNZ 644 (CA), applied in Iongi at [107].

The defence does not need to break every strand. Doubt frays a strand. So does a reasonable alternative the prosecution has not closed off. Fray enough strands and the rope cannot hold.

The prosecution weaves the strands into a rope. The defence tests each strand, and the alternatives the prosecution has not closed off.