Evidence Course Home Page

 

---

Evidence Course
Professor Peter Tillers Cardozo Law School

---

 

A Benthamite, Baconian, or Toulminesque Account of Inference

We have considered one approach – Bayesianism – to factual inference. Now let’s consider an approach that may be a bit different. This an approach that the editors of your coursebook discuss at pp. 84-91. What appears here is a restatement of their account.

The law of evidence is concerned with inference, the drawing of conclusions about the world from evidence.

To describe how inference works, the law of evidence, both in the United States and in the Commonwealth, regularly makes use of terminology used by Jeremy Bentham.

Figure 1

 

Bentham referred to the fact to be proven as the

factum probandum

He referred to the matter that may tend to prove or show some fact, some factum probandum, as the

factum probans

the fact or matter that does the proving.

The step from the “factum probans” to the “factum probandum” is

inference

This terminology is still used.

&&&

Please note two important things about Bentham’s account of factual inference and proof:

First, factual inference is seen as a kind of rational process. Bentham applied the language of logic – “inference” – to factual inference and proof. Bentham fully understood that factual inference is different from deductive inference. But both he and American legal discourse treat factual inference as a form of reasoning or argument in which some proposition or ingredient -- the factum probans -- serves as the premise, some other proposition -- the factum probandum -- serves as the conclusion, and the word inference is used to describe the step from the premise to the conclusion.

Second, Bentham’s terminology suggests that factual inference does not always or possibly ever involve the drawing of a conclusion directly from evidence to a factual conclusion. Bentham’s terminology refers to the premise of evidential inference as a “factum probans.” In modern terminology, the factum probans is a factual proposition; i.e., the matter that serves as a premise for a factual inference is a proposition rather than a thing. This way of talking suggests that there is no “direct” apprehension of “evidence.” It suggests, among other things, that different eyes and minds see the same evidence and the same sensory phenomena differently.

Some older American judicial opinions sometimes say – and even some modern opinions sometimes say – that (i) the drawing of inferences is a property only of circumstantial evidence and that (ii) real and tangible evidence are different because in those situations the thing itself -- or direct evidence of some matter in issue -- is presented to the trier of fact and no inference is involved or necessary.

Today, however, most judges, logicians, legal theorists, probability theorists, and students of the philosophy of science disagree with this attempted distinction between circumstantial and “real evidence” or “direct evidence.” Most students of inference today think there is no such thing as direct evidence and that the drawing of conclusions from evidence always involves inference, uncertain inference.

Examples of direct apprehension of facts evaporate on close analysis. Thus, what appears to be a spot of blood may be a catsup stain, what appears to be a knife wound may be a gunshot wound, and what appears to be the sun may be a reflection of sunlight -- and, in any event, no one ever sees anything directly since all “seeing” is mediated by the brain, sensory organs, and so on -- and the same holds true for sounds and smells and touch and so on.

It is true that some inferences seem very, very probable -- some inferences practically certain -- but this does not derogate from the fact that practically-indubitable inferences are still inferences.

&&&

But the picture thus far presented of inference may be incomplete. Some observers think that an account of rational inference must explain why an inference -- any inference -- is drawn from some evidence: some observers think that there must be some warrant for a factual inference. These observers typically say that the required warrant is some "generalization"; they sometimes say that generalizations are the "glue" that tie evidential premises to inferred factual conclusions.

Consider an example. We will shortly study cases in which E escape from the clutches of the law is offered as evidence of someone's G guilt of some crime. The question arises: How or why is escape evidence that someone is guilty of some crime? The problem we face is perhaps this:

Figure 3

&&&

The above visual representations of factual inference are, of course, too simple. But is it possible that the figures found above contain all of the elements -- all of the building blocks -- that one needs to construct more intricate explanations or representations of more complex inferential reasoning?