Seeing is Believing Part 2 – On Plausibility

“Seeing is believing” – or is it?

Remember from Part 1, published a month ago, that my venerable Aunt Ethel claimed, “On my birthday I saw Humpty-Dumpty take a big fall.”

She is not blind, rarely wears glasses, hears quite well, so what did she see? Since none of the witness believe that a person or object corresponding to the description of Humpty-Dumpty exists in the sense that it can be seen wandering around the neighbourhood in the ”real” palpable world, Aunt Ethel’s perception may have played her a trick.

The obvious solution is that she mistook a rather small, round, egg-shaped, bald gentleman ambling around the cricket field as Humpty-Dumpty. Such mis-identifications are common enough. Or perhaps she omitted a word or two and wanted to say, “I saw someone who reminded me of…”

She surely did not mean that there is more than one Humpty-Dumpty since the name is not a tag for a class which has more than one member, but refers to a class which has a defining set of properties which limits it to one member only.

Therefore it is unlike Tweedledum and Tweedledee, which is a class specifically limited to two members. However, the class is defined in a manner which precludes their instantiation. In common parlance, they are fictional characters which have many attributes — in fact, more attributes than may be found in some catalog of attributes but which include that these do not have life-signs, like “amble about of their own free will!” Such attributes specifically preclude that they could be encountered on the street, on a boat, or in an airplane — that is, in living form. We often describe something by stating explicitly what properties it does not have, e.g., an instance of vandalism as being, in part, thoughtless acts, compared to those which were instances of willful destruction.

The general principle is this: There are words and expressions which we use to report to others:

  1. that we have made an encounter of some kind
  2. that what was encountered can be described to others so that others can identify the object of our encounter on the basis of some limited selection of items which we have selected from a larger set than was provided, and
  3. that the identification of the object of the encounter can be facilitated by assigning to it a tag (perhaps a name, like Humpty-Dumpty) by which it then becomes more widely known.

The tag that refers to a class of one only (like Emperor Napoleon I or Emperor Napoleon III or Picasso’s portrait of Gertrude Stein) is all that is needed to enable others to identify the object of my encounter. But what is involved is that the report of my encounter involves a hypothesis to the effect that the encounter has been coded by myself in the hope that it can be readily decoded by others without them making a critical error, a fatal mistake in identification. Since coding means that only some of the attributes of an encounter are contained in the message, the probability of an error in recognition by others is relatively high.

To illustrate this, consider the case where the Mafia boss says “get me Joe Bloomfield” and his henchmen shoot down the hamburger vendor on 10th Street, not the barber on 42nd at Times Square. What the two men have in common is a name, their location in New York City and some other qualities, e.g. gender. So the name served to locate the object or, in this unfortunate case, the target, but the name did not inform whether sufficient attributes were listed to sort it from a mass of other potential targets.

The key issue for discussion is to sort out claims made by others — whether it is the claim made by my Aunt Ethel or by a witness in a court of law — about the relation between what they claim to have witnessed and the plausibility of their testimony. In the case of Aunt Ethel the plausibility of her claim is zero because she failed to acknowledge that she was a witness to “a likeness of” and not a witness to an “encounter with the one and only fictional character described by Lewis Carroll named Humpty-Dumpty.” It is a common enough error made in reporting, but in this case it stands out because the claim is utterly implausible and by inference, also improbable. Aunt Ethel could not have seen what she claimed to have witnessed on the ground of implausibility.

Seeing is Believing — Or Is It?

You don’t see a Belief, but believe what you see — but should you?

“Seeing is believing” — or is it?

“You see one — of many?”

“On my birthday I saw Humpty-Dumpty take a big fall,” says Aunt Ethel, who turned 102 yesterday.

What did she see?

She is not blind, rarely wears glasses, hears quite well. She is acute and astute. We assume that she must have seen something, but did she see what she reported? Since none of her interlocutors believe that a person or object corresponding to the description of Humpty-Dumpty exists in the “real”, palpable world — except in a much-loved story — Aunt Ethel’s perception played her a trick. Or so we conclude.

But if we are wrong, and Aunt Ethel’s sighting can be confirmed, would she be entitled to claim, “I saw a Humpty-Dumpty. Furthermore, there are more Humpty-Dumptys somewhere, in fact, I saw only one of many.”

We are all Aunt Ethels in this respect because we all tend to assume on the basis of one sighting that there are more of the same, unless the sighting has a particularizing, individuating name, or tag. The logic runs: “a class of one is still a class” — call it a sample of a class.

I saw a black swan 10 days ago — ergo, there are black swans, or black swans exist. Not only one, but some (many).

To say “I saw the only black swan 10 days ago” is a very strong hypothesis which runs as follows: “Normally there are no black swans, but I saw one and since I believe what I see, therefore I also believe that black swans exist, but perhaps only one!”

What I have done is to clarify what is meant by holding a belief – by assuming that something which is against the grain can nevertheless be incorporated with a set of expectations which I have come to share – or have unconsciously learned to share — with others.

My Aunt Ethel simply followed a widely-adopted convention when she suggested that seeing a one of something is an acceptable basis for assuming that uniqueness is not “normal” at all, but that a single case is a good and fair ground for assuming that there are more of the same somewhere, some place. Keep counting.