Levels of knowing: A scale for assessing uncertain assertions
How to evaluate how confident you can be about an
assertions
by Tom Cloyd, MS, MA – Counselor / Psychotherapist – Bellingham, Washington (360) 920-1226 – email: tc (AT) tomcloyd.com (please read about content licensing)
Telling the truth is a skill. Those who don’t do it habitually lose
the ability… ~ David Brooks
Truth is the safest lie. ~ Yiddish proverb
STATUS OF THIS ARTICLE: This is an incomplete first draft. Much of the core content is here, but some of the ideas are seriously undeveloped or underdeveloped. It’s made available in its present state anyway because its ideas, which are of fundamental importance, are referred to in other articles in the Library.
SUMMARY
“Knowing” isn’t an either/or thing. There are degrees of knowing. Virtually nothing is known for sure, but many things are known with something less than complete confidence. Use of the “Levels of Knowing” scale provided here can tell you approximately how MUCH you know, and keep you from going beyond the limits of your knowledge.
Introduction: What this is, and why it’s here
Can you trust what you think you know? Can you trust what you think someone else knows – or what they say they know? This is one of the great problems of life. Many people do not handle this problem well, but then it’s not an easy problem.
When you ACT AS IF SOMETHING IS TRUE, this indicates that you feel a certain degree of confidence that it IS. Whether or not something is true is a question that we face all the time. Sometimes the question is merely interesting, but other times it is a very critical concern.
Problems arise mainly when (1) we need to know something that we don’t yet quite know, or when (2) we act as if we confidently know something that in truth is fairly uncertain. Knowing the difference can be critical.
“Levels of certainty” table
The purpose of the table below is to clarify the degrees of certainty we may have concerning what are known in philosophy as “propositions” or “assertions” – namely, statements that “A” is “B”. Here are some example assertions:
- Albert is an alligator.
- My refrigerator is broken.
- The end of the world is near.
The task we face with any assertion is to evaluate how much confidence we might properly have that it is true. This scale, contained in the table below, can help you clarify your proper level of confidence.
| Assertion type (examples in parentheses) | Explanation | Appropriate degree of certainty that the assertion is true |
LEVEL 1: concept of a possibility (e.g., “perhaps this orange was put here by space aliens”) |
This is just an idea, with essentially NO empirical or experiential basis for belief. “Brainstorming” sessions products lots of these sorts of assertions. |
None at all. |
LEVEL 2: hunch (e.g., “I think this orange was left here by a customer”) |
The assertion has a degree of plausibility, in that it could be true. If it were true, then the event of interest would be accounted for. However concrete evidence for its truth is lacking. |
None, although one may feel fairly certain that the assertion cannot yet be considered untrue. |
LEVEL 3: belief (e.g., “Surely this orange was left here by a customer”) |
This is anything from a concept of a possibility to a hunch to an anecdotal hypothesis, to which an unjustified degree of felt certainty has been attached – enough that the believer is willing to use the belief as the basis of action. Beliefs survive as such because they are correct, or because the consequences of their being in error are minor or not directly or immediately experienced. |
From none to minimal, depending upon the formal nature of the assertion underlying the belief (see previous column). |
LEVEL 4: hypothesis, casual |
This is a assertion, put in language that allows testing of its truth value, which has a basis in the (presumably correctly) reported experience of experience of one or more people. |
Minimal. The problem is invariably one of sampling: the experience reported is not demonstrably representative of the actual state of affairs in domain in which an interest is taken. |
LEVEL 5: hypothesis, formal |
The assertion has a basis in anecdotal report, theory, or one or more formally conducted research investigations. |
Minimal to moderate, depending upon the degree of support given by the supporting research and degree of doubt suggested by research reaching a contrary conclusion. |
LEVEL 6: research finding (example to be provided) |
The assertion has a basis in supportive results of one or more formally conducted research investigations. |
Minimal to moderate (but never as low as the highest ranking research hypothesis), depending upon the degree of support given by the supporting research and degree of doubt suggested by research reaching a contrary conclusion. |
LEVEL 7: theory (example to be provided) |
This is an integrated set of hypotheses, findings, and theories, which address broad ranges of phenomena. |
Moderate, depending upon the empirical bases of the theory, and the degree of internal coherence and external explanative power it possesses. |
LEVEL 8: model, provisional ( |
This is the same as a theory, but its terms are better specified – more specific, more measurable, more capable of generating specific predictions. |
Moderate, depending upon the empirical bases of the model, and the degree of internal coherence and specific external explanative power it possesses. |
Level 9: model, working (empirical – example |
This is the same as a provisional model, but is even better specified, usually affording a high degree of measurability and specificity of outcome prediction. |
High. No theory becomes a working model without having had to prove itself to be robust in the real world. |
How to print this page
Use the page print function built into your browser. It's usually an item in the File menu right below the caption of the window in which this page is displayed.
Only the page's core content will be printed. All images outside of the main content area, all navigation tools and links, and extraneous header and footer material will be omitted from the printed page.
Many browsers also have a print preview function on the same menu - you can use this to see how the printed page will look before you actually print it.