Haah! You believe that? Irrational much!

Why rationality does not guarantee true beliefs.

You post an elaborate argument in favor of or against a belief you hold or contest, respectively. Ideally, detractors engaging with the post would provide a clear explanation for their disagreement — be it a logical error or the usage of some insufficiently substantiated premise. Unfortunately, some random person comments something on the lines of “Haah! You believe that? Irrational much!”, referring to the conclusion rather than the reasoning to deduce irrationality.

Haah! You believe that? Irrational much!

The usage of “irrational” with the implication of diminished veracity indicates characterizing rationality as being related to truth. In particular, it characterizes the lack of rationality (irrationality) as entailing false beliefs, and possibly the presence of it as entailing true ones.

The present post argues that these characterizations of “rationality” are inherently either contradictory or useless depending on the context, as well as teasing at my argument for the conception of rationality as the affirmation of justified beliefs, which includes the affirmation of uncertainty claims when justified, as well as the rejection of negative ones.

Truth-Justification Non-Equivalence

What is deemed knowledge undergoes constant changes. Up until the 16th century, the geocentric model — the astronomical model positioning the Earth as the "center of the universe” — was a prevalent dogma. The sun, planets and other stars appeared to revolve around the apparently stationary and solid Earth marking each day. Given these observations, geocentrism seemed to be the inference to the best explanation. Affirmation of inferences to the best explanation is justified given the lack of better explanations (by definition), so long as the provisional and uncertain nature of the inference is acknowledged; for, an inference is not a deduction (should it have been a deduction, it would cease to provide new empirical information).

Nevertheless, as virtually everyone today acknowledges, geocentrism is simply false. It seems that this once-justified belief was false.

Not all that is justified is true.

At the same time, to consider a model of heliocentrism should have seemed unprompted: isn't the sun just the day version of the moon? What's so special about it?

For many cultures, then, heliocentrism would not have been a justified position lest additional information is provided — as conveyed in the 16th century by Galileo, Kepler, and Corpenicus. Perhaps even a century following, many of our current ideas are to be superceded by more accurate ones, even if those ideas provide only a keyhole to at least a glimpse of reality.

Not all that is true is justified

As counterintuitive as this pair of mutually converse statements might appear, their veracity reigns lucid upon further reflection. Justification concerns the integrity of our reasoning processes, not the certainty of their outcomes.

Justification-Affirmation Non-Equivalence

The beliefs a typical individual holds often do not align with the beliefs they should hold. The prescribed beliefs must maintain consistency not only with other beliefs held, but with the results of truth-seeking methods they deem reliable and notably satisfactory, as well.

In detriment to that proposition, confirmation bias yields only considering information consistent with our existing beliefs, even if they contradict the results of methods we claim to affirm in other contexts. Availability heuristic infleunces greater weight paid to information easily accessible in memory, which can lead to affirmation of claims inconsistent with one's otherwise espoused methods, as can do cognitive biases more generally. Logical fallacies both formal and informal are prevalent and pervade virtually all of human discourse. Beyond informal fallacies hindering both of the described forms of consistency, formal fallacies compromise internal consistency by definition. Clearly, the justified (or not so) is not equivalent to what we affirm (or reject).

Not all that is affirmed is justified

Not all that is justified is affirmed

Truth-Affirmation Non-Equivalence

Truth-affirmation equivalence would entail everyone agreeing on a single truth in a matter, which, evidently, is unattainable.

Useless Conceptions of Rationality

The pairwise non-equivalences of truth, affirmation and justification would entail that if rationality is determined based on only one of them, it cannot be determined through only the two other variables. Therefore, to demonstrate that rationality does not guarantee true beliefs, one may show that it cannot be characterized as affirming only justified true beliefs.

Rationality and Truth

Let us return to the initiating quote:

Haah! You believe that? Irrational much!

It was noted that this quote implies rationality to entail affirming true beliefs, presumably justified ones. Nevertheless, we have seen that de facto affirmation, justification and truth are not really equivalent; and that while justification can point to the truth, it cannot provide us the truth. One may be more charitable and steel-man the quote by seeking an interpretation which does not assume the disproven equivalences: perhaps the quote merely uses a different conception of rationality.

It is clear that, irrespective of whether it implies affirmation and justification to be necessarily consistent with truth, it implies affirming true beliefs. However, this conception diminishes the systematic nature of rationality, unless one posits the existence of an absolutely certain means of knowledge acquisition. The shift away from geocentrism, from the conception of spacetime as absolute to general relativity and from a creationist view of life to evolution all demonstrate the uncertainty in once mainstream epistemic affirmations, and we continue to discover more about the world every day. It, therefore, is clear that for rationality to be systematic, it cannot be the affirmation of truth per se.

The only other possibility is the affirmation of the justified - consistent with the connotation of rationality being systematic.

Rationality and Justification

I will be steel-manning the quote further. Perhaps, it does conceptualize rationality as affirming justified beliefs, but this yields another problem. In the context of merely having heard of or read someone’s claim without their justifications, to reject it as not justified, unless the claim is conspicuously self-contradictory, is to merely assume their irrationality.

Alternatively, one may even conceptualize justification differently. Perhaps, for the quote, whether a belief is justified or not is a property of the belief itself. For instance, that “I am eating a carrot and I am not eating a carrot” under typical interpretations is conspicuously not justified given its obviously contradictory nature. Under this interpretation, should we observe new data strongly suggesting that the topology of the universe is that of a donut (torus), and that belief becomes justified, then it was always justified and alternatives to it were never justified (assuming consistency: that contradictory claims cannot be simultaneously justified). This would mean that we cannot even ascertain justification, let alone truth. Since rationality is assumed to be an affirmation of justified beliefs, this predicament compromises the systematic nature of rationality as well.

It is clear that, for a systematic conception of rationality, we require not only that it is the affirmation of justified beliefs irrespective of absolute truth, but also that justification is not a property of truth-claims per se.

What if the quote refers to the truth or justification of the premises (than conclusion) of the allegedly irrational?

If the conclusion and/or belief is the outcome of the process of reasoning, then the initial conditions or the variables determining the process are the premises one uses. In the imagined context, however, the accusing interlocutor does not consider the premises, and refers to the conclusion. Thus, this interpretation is tautologically ruled out based on the context if not the quote itself. However, I will later argue that, this too, contradicts the systematic nature of rationality.

Summary

In this section, I attempted an extensive steel-manning of an imagined accusation of irrationality based solely an observation of a claim. Based on the inference that the accusation is more specifically based on a variable related to the truth of the claim, I ruled out the possibility of that variable being the truth per se as well as “justification” if framed as an intensive property. Furthermore, the extension the set of truth (of the conclusion) and an intensive notion of justification ({truth, “justification”}) as variables to contain the premises ({truth, “justification”, premises}) ensures that all analogous interpretations of the accusation quote are considered (under the assumption that all the variables related to truth concern either the process, initial conditions or the outcome — as they determine the entire epistemic landscape in this context). Given the presence of at least some contradiction given every interpretation of the quote, especially with the implied systematic nature of rationality, it can safely be considered invalid.

The Acceptable and Accepted Notions of Rationality (?)

The analysis in the prior section implies that for rationality to be systematic, it must be a function of justification. We further eliminated the possibility of justification being an intensive property of claims (i.e. that each claim in itself is rational or irrational). Since rationality is effectively an affirmation of justified belief, to characterize “justification” in context is to characterize rationality.

While a specific instance of epistemic rationality may not be instrumentally rational and vice versa, this characterization of rationality as a whole as the affirmation of the justified (epistemic rationality) gives away to the characterization of rationality as the alignment of decisions with goals (instrumental rationality). To see how, associate to each decision the claim “the decision is aligned with my goals” - which entails instrumental rationality to be the affirmation of justified claims of the form “decision X is aligned with my goals”.

As for my exact characterization of justification itself, it will be explored in the next installment of Abstreal Logoce.

Conclusion

This exploration has laid the foundation for a systematic conception of rationality, one emphasizing the alignment of beliefs and decisions with justification rather than an a non-existent equivalence to truth. Through the dissection of relationships between truth, justification and affirmation, as well as comprehensive steel-manning, we established that rationality cannot be characterized as merely the affirmation of true beliefs. Instead, it is the process of grounding beliefs in justification irrespective of their (elusive) ultimate veracity, especially due to the expected systematic nature of rationality.