I had this wonderful idea for an app that would use bivalent logic to evaluate and flag fake news.
Although bivalent or “missing middle” logic was invented by Aristotle (384 – 322 BC) in Organon (see the square of oppositions) and can be thought of as the basis for the microchip, where an electrical impulse can follow a circuit through different paths such as AND, IF (if and only if), OR or NOT gates (since replaced by the concept of negation as failure and yet to be replaced by quantum superposition), the main attraction for this purpose, the evaluation of text, is that it converts directly to binary code. Wa-hey! Low hanging fruit. Digital evaluation can produce a quantified output, which can then generate a probability of (for example) bias, hoax or even a fake review, without being clouded by the interpretation of the observer.
This is how bias can be revealed in text:
Fake claims can be identified by frequency of first tier unproven statements and instances of the reinforcement of first tier claims citing second tier unproven statements as factual. A fake review can be identified by an over-use of superlatives and low proportion of negative criticism.
Something either is or it isn’t; 0 or 1, so every statement has one of two possible truth values, true (1) or the negative of truth (0). Almost nothing non-mathematical can be proven to be true so the benchmark for this has to be what the accepted literature takes as “proven” to the current extent of our knowledge at this time. UFOs, for example, are proven – anything positioned in the air that you can’t quite identify. Aliens are not, so claiming a UFO = 1 but claiming it is therefore alien = 0. It cannot be both true and false, so that’s a bad mark for the trustworthiness of the text (unless it becomes the accepted reality later, when the literature catches up).
Statement: We build. Non-statement: Should we build? Opinion: We should not build. Command: Build! The truth value is assigned based on whether it happened or not, i.e. did we build (1) or not(0)? The assessment of whether something is a statement or a non-statement is important.
(With non-contradiction) that leaves seven sets of binary truth values shown as digits:
A (it is) = T, T, F, F = 1100
B = T, F, T, F = 1010
¬A (it is not) = F, F, T, T – 0111
¬B = F, T, F, T = 0101
A ∧¬A = T, T, T, T = 0000
AV¬A (it is or it is not) = T, T, T, T = 1111
A∧B – T, F, F, F = 1000
To give you an idea without the numbers, lots of praise in a review and hardly any destructive questioning would leave a tally like 19 x F, 1 x T = 95% chance it has been written by a friend of the author. A high proportion of unfactual or unsubstantiated statements in an article would leave a similar trail, question mark count, for a fake or biased news report. A significant count of rubbishing the alternative would also flag as propaganda. This would initially need a human assessor until we can connect a database or encyclopaedia to match claims against. Results for tabloid newspaper editorials or scores for the reliability of individual journalists would be very revealing.
The app could be expanded to spot ideological leanings or favouritism being expressed by the writer for one nation or race over another. To clarify, using a particular word would not generate a result, e.g. “the foundation studied communism/fascism/racism for years”, but when the subject has been frequently referred to in conjunction with words such as “like”, “love” or saying the subject is “the best system” or “obviously correct” those would be identified, as would denigrating or stigmatising words against the subject’s diametric opposition. The only way a propaganda writer could avoid having their article flagged as propaganda and themselves as biased would be to take out the favouritism and stick to selective recounting of the facts, which would be a big improvement.
How should I rate this book? In mathematics, the result is either perfectly correct or it isn’t and in bivalence it is or it is not, so this book is either correct or it isn’t and I could give it 5 stars but that would be a world gone topsy turvy when there’s an unnecessary “of” in the Descartes section toward the end and also I’m a subjective human reviewer rather than a robot, so I’ve unilaterally deducted a star from this book because it gave me a headache. Clearly, I’m not bivalent. Byte me.