email - September 2016
by Do-While Jones

Scientific Inference

When is inference valid?

Joseph had this comment about last month’s article on Arbitrary Classification:

My only statement is that the word "infer" means to "draw a conclusion based upon evidence." The word is not inherently supposed to mean, "To just guess something, as best you can tell."

I feel like you tend to see the word as the latter definition, although I admit that neither definition makes the word have a strong scientific slant.

Joseph brings up an excellent point about the definition of infer, and its two possible meanings. Let’s compare his two different informal definitions with Webster’s.

Webster’s Definitions

Simple Definition of infer
: to form (an opinion) from evidence : to reach (a conclusion) based on known facts
: to hint or suggest (something)

Full Definition of infer
1 :  to derive as a conclusion from facts or premises (we see smoke and infer fire — L. A. White) — compare imply
2 :  guess, surmise (your letter … allows me to infer that you are as well as ever — O. W. Holmes †1935)
3a :  to involve as a normal outcome of thought
3b :  to point out :  indicate (this doth infer the zeal I had to see him — Shakespeare)(another survey…infers that two-thirds of all present computer installations are not paying for themselves — H. R. Chellman)
4 :  suggest, hint (are you inferring I'm in­competent? ) 1

The dictionary says both definitions are valid; but rather than quibble about the legitimacy of the definition, let’s focus on the important point that Joseph makes—specifically that inferences can be valid, reasonable conclusions, not simply wild guesses.

The National Science Teachers Association (NSTA) says:

In science a theory is a well-substantiated explanation of some aspect of the natural world that can incorporate facts, laws, inferences, and tested hypotheses (NAS 1998). 2

In light of the fact that inferences may be valid, or may simply be speculation, what place do inferences have in science? How do we tell when an inference is a reasonable conclusion based on facts, and when is it just a wild guess?

Where There’s Smoke

By today (August 22, 2016) the Blue Cut Fire (which began 90 miles south of here on August 16, and had closed Interstate 15 for two days) had burned 37,020 acres and was 89% contained. The Cedar Fire (50 miles west of here) had burned 19,214 acres and was just 5% contained. The combined smoke from both fires was so thick that our private school closed today out of concern for the health of our students. Certainly there was lots of smoke here—but no fire here. Not a single flame. Therefore, one might try to argue that the inference, “Where there is smoke, there is fire,” is not always valid. The inference is incorrectly applied if the word, “where,” is taken to be too geographically specific. If “where” is taken to mean, “southern California,” instead of “Ridgecrest,” it is true that where there is smoke there is fire.

It’s dangerous to draw conclusions from inferences because the inference might be incorrectly applied. It might be true in some cases, but not others.


When I walk through the desert, I sometimes notice tracks in the sand, and I can infer that a coyote (not a motorcycle) made the tracks. There is no possibility that I might mistake coyote tracks for motorcycle tracks even though I didn’t see the tracks being made. I can make a correct inference because I have seen coyotes make a trail of paw prints in the past. I have seen motorcycles leave tire tracks. I’ve seen the different causes, and I’ve seen their effects. When I see a familiar effect, I can infer the cause.

On those rare occasions when it rains here, I’ve seen raindrops make marks in the sand. They are randomly spaced, randomly sized depressions. The probability that raindrops will make depressions in the sand just like a coyote paw is very, very small. But, given enough trials, very improbable things can happen. Given the exceedingly large number of drops in just one rainstorm, theoretically, it is remotely possible that large raindrops could land in a pattern identical to the pattern produced by the paw of a coyote. And, if it could happen once, it could happen multiple times. So, if there is a series of depressions in the sand that form a line in a particular direction, theoretically, that trail could have been caused by the rain—but I’ve never seen it happen. I’ve never read any report that says other people have seen it happen. So, when seeing a trail of depressions in the sand that look just like the paw prints coyotes make, it is reasonable to infer that a coyote made them. It isn’t reasonable to infer that the trail was made by the rain. And it certainly isn’t reasonable to assume the trail was made by a motorcycle.

If you have observed the cause and its effect many times, and you have never seen the effect produced by any other cause, you can legitimately infer the cause from the effect.

Inferring Motive

So, we see a trail of coyote footprints. Why did the coyote cross the sand? Sadly, we don’t have a coyote crossing the sand punch line. Some might say he was looking for a rabbit to eat—but we can’t absolutely infer motive from the tracks. The tracks might be distorted in the way they get distorted when a coyote is running rather than walking; but that still doesn’t mean the coyote was chasing a rabbit. If the tracks are very clean, it might mean the coyote was sneaking up on a rabbit rather than running after it. But if there aren’t any rabbit tracks around, how do we even know there was a rabbit in the area? And if there are rabbit tracks, how do we know they were made at the same time? Maybe the rabbit hopped along here after the coyote was gone.

The Watchmaker Inference

In 1802, William Paley argued that if a pocket watch is found on a heath, it is most reasonable to assume that someone dropped it and that it was made by one or more watchmakers, and not by natural forces. We’ve all seen pocket watches, wrist watches, and grandfather clocks. Every single one of them was made by a watchmaker. Nobody has ever seen a pocket watch that was not made by a watchmaker. Therefore, it is reasonable to infer the existence of a watchmaker from the discovery of a watch. But it is not possible to infer who that watchmaker was, or why he made it.

Evolutionary Inferences

Now that we have clarified our agreement with Joseph that inferences aren’t always simply wild guesses, and can be part of a legitimate scientific theory, let’s see if evolutionary inferences are legitimate.

Dawkins says that one can infer evolution (but one cannnot infer the existence of an optical designer) simply from the existence of eyes.

It has been authoritatively estimated that eyes have evolved no fewer than forty times, and probably more than sixty times, independently in various parts of the animal kingdom. In some cases these eyes use radically different principles. Nine distinct principles have been recognized among the forty to sixty independently evolved eyes. I'll mention some of the nine basic eye types-which we can think of as nine distinct peaks in different parts of Mount Improbable's massif-as I go on. 3

How many documented cases of optical systems designed on purpose are there? Very many. How many times have optical systems been observed to have been caused by random chance? Not once. Why should one infer a cause (evolution) that has never been observed instead of a cause (design) that is commonly observed?

Apparently evolutionists believe inference is a valid form of reasoning when it is unfounded speculation that is consistent with their prejudice, but not when it is a reasonable conclusion based on observable evidence if it contradicts their prejudice.

Has anyone ever observed a sweat gland evolving into a mammory gland? No, they haven’t. That’s why we rejected the inference that mammory glands evolved from sweat glands. 4

Evolutionists infer motive. If you haven’t heard an evolutionist make up some fairy tale about why some characteristic evolved, just pay attention. You soon will.

Whenever you come across any evolutionary theory about how something evolved, you should ask yourself, “Is this inference based on an established track record of observance, or is it just wild speculation?”

Quick links to
Science Against Evolution
Home Page
Back issues of
(our newsletter)
Web Site
of the Month
Topical Index


3 Dawkins, Climbing Mount Improbable, pages 139 - 140
4 Disclosure, January 2002, “Sweating Milk”