As part of Sam Harris’ Waking Up podcast, Harris invited someone I had not heard of ― Jordan B Peterson ― with a schedule that, I presume, included discussions of morality and religion. However, Harris and Peterson ended up having a disagreement about what it means for something to be true. Harris didn’t really offer a definition of truth, and so I can only fill in a guessed working definition for Harris. Peterson, however, did give a definition of truth. I want to criticise Peterson’s definition and offer some of my own thoughts.
Peterson defined his conception of truth as both Darwinian and pragmatic. What this means is this: a thing is more true the more useful it is, and the metric of usefulness is ― quite literally ― evolutionary success; if a claim hinders human biological progress, then it is ‘not sufficiently true’. Harris challenges this view by imagining two virology labs, both of which have identical and accurate empirical conceptions of the smallpox virus. One of them uses this knowledge to create a cure, whereas the other uses this knowledge to manufacture a weapon. Under Peterson’s conception of truth, the cure-creating lab has a ‘truer’ conception of the smallpox virus than the “nefarious” lab. This is despite the labs having identical conceptions.
if it doesn’t serve life, it’s not true.
– Jordan B Peterson
Peterson defends this position ― the position whereby the nefarious lab’s conception of the smallpox virus is ‘not sufficiently true’ ― by claiming that moral truths are a deeper truth than scientific or empirical truth. And because the nefarious lab has failed on some moral grounds, then similarly it fails on truth grounds. The ‘moral ground’ is equally grounded within the idea literally Darwinian success. Peterson ties the intentions of the nefarious lab with the claims they understand. The ‘truth’ we don’t have here, which apparently corrupts the whole enterprise, is the truth on how to modulate the behaviour of nefarious people.
Harris goes a step further, all the while finding synonyms like ‘accurate’ and ‘description of reality’, and talks of two labs with identical conceptions of smallpox (‘true to a first approximation’, and good enough to synthesis either cures or weapons from) and equally good intentions. However, one lab ― completely by accident ― has a small breach and releases smallpox into the general population, creating a horrific epidemic. Under this circumstance, is one lab further from the truth about smallpox than the other? Again, Peterson argues that the answer is ‘yes’; this time the truth has been corrupted by a clearly failed discussion about the ethics of science: the tool is ‘pretty damned dangerous’ and thus the research should not have been happening. It is moral truths in that discussion we do not have this time, and that corrupts the whole enterprise.
Peterson’s model is this: if a description of reality is destructive in any sense, then there has been some pragmatic failure. That failure, in turn, corrupts the description of reality (despite any claims of empirical accuracy).
The claim I’m making is that scientific truth is nested inside moral truth, and moral truth is the final adjudicator. And your claim is ‘No! Moral truth is nested within scientific truth and scientific truth is the final adjudicator’.
– Jordan B Peterson
I do have a number of areas of agreement with Peterson: I do think ‘truth’ has Darwinian elements to it, but not in such a literal sense; and I do think there are pragmatic concerns in ‘truth’. However, I also disagree: where Peterson thinks the hierarchy of truth has ‘moral truth’ at its deepest level and scientific truth is subservient to that, I think ‘reality’ is the adjudicator of truth and scientific truth is our best approximation to that truth (better called ‘knowledge’ than ‘truth’), and moral truths emerge from there. (I have a scheduled post in April that argues moral truths are parochial. Still true, but not universal.)
First, the agreements: truth is Darwinian. Peterson thinks truth is literally Darwinian, where knowledge is a part of the ‘extended phenotype’, and the ability of that phenotype to propagate is the arbiter of truth. I think knowledge is figuratively Darwinian, competing is a very different environment. Knowledge competes in a rational environment, where success is measured in citations and failure is simply falling out of being cited. Descriptions of reality are tweaked (mutated) and if that mutation allows it to better account for reality it is cited more (selection). Worse mutations are measured by the vigour with which they are hit in peer-review. This is what Deutsch calls a ‘rational meme’. By contrast, ‘nonrational memes’ are claims that spread by appealing to nonrational parts of the mind or aggressively attacking and demonising competing views.
The truth value of a proposition can be evaluated whether or not this is a fact worth knowing or whether or not it’s dangerous to know.
– Sam Harris
I also think there are pragmatic judgements to make in the issue of what is true. In this sense, like Peterson, I am a pragmatist. But it is not an evolutionary pragmatism, like the one Peterson espoused. Instead, my pragmatic judgement is one where we carefully ring fence off different topics. This is in stark contrast to Peterson’s conception. Where Peterson intentionally ring fences truth claims with moral considerations, thus groups an accurate conception of smallpox with the nefarious intentions of a person who understands this conception, I believe these are very different topics. Hypothetically, if a terrorist organisation gets a hold of a stockpile of nuclear weapons and places a weapon in every city and town on the planet and then detonates them at the same time, Peterson would argue this act works to demonstrate our understanding of nuclear physics is ‘not sufficiently true’; I very strongly disagree: the detonation is further empirical evidence of the truth of our understanding of nuclear physics; what we were wrong about is what constitutes ‘adequate nuclear security’.
There is a conversation to be had about which domain is more important to society. Moral considerations are more important when talking about how a society should organise itself, and funding scientific research that is clearly dangerous may not be sensible. But that priority in social structure is not indicative of ‘deeper truths’. There may be a case where we decide not to properly investigate the differences between human ‘races’, because if we find a difference in intelligence or humour or sociability (or anything else we tend to value) that may be destructive to our society. But that appears, to me, to be an admission that the scientific truth would be true regardless of our moral wishes; if anything, we know it would be true, regardless of its moral implications.
Many commentators, and even Harris, has noted that this conversation seems like a distraction from what they presumably scheduled to talk about: religion and morality. As I see it, though, it is completely necessary ground work. Imagine a conversation between Harris and Peterson talk about religion from within their differing views of truth. Peterson could easily claim religions are ‘true’ because they are useful, even if they do not comport with reality. If a religion aids social structures and therefore survival, it becomes both true and moral, in Peterson’s view. This is independent of whether the two religions are contradictory, and independent of whether the religions increase wellbeing or comport to reality. The apologists old dodge of ‘religion is useful’ would be given a whole new lease of life: ‘religion is true, because it’s useful’.
I don’t know Peterson’s view on the topic of religion, and I’m sure his book The Architecture of Belief will make for very interesting reading. But a concept of ‘truth’ that unties the claim from reality, by applying a moral judgement to it, allows him to claim just about anything is true ― so long as a good person knows it.