Via formspring: Kant and a priori synthetic judgements – why are they necessary for the possibility of experience?
Currently in my formspring inbox (along with two other questions I’ll answer in due course)…
“Kant claims that a priori synthetic judgments are necessary for the possibility of experience. What does he mean by this?”
Here we go!
Some terms and some history
Prior to Kant’s arrival on the Enlightenment scene, judgements were divided into two categories – a priori and a posteriori. A priori judgements were made via pure reason (it is an a priori truth that A=A), whereas a posteriori judgements are derived from experience, therefore being unreliable in their specific application (it is an a posteriori truth that some swans are white – but some aren’t, as you can see in my title bar above).
Put as an oversimplification the Enlightenment rationalists were all about a priori judgements and the Enlightenment empiricists were all about a posteriori. (Which is why I get a bit miffed when people who are all about the empirical evidence call themselves “rationalists” – but that’s another story*). Suffice to say, this left the empiricists with all the practical application (i.e. evidence based), while the rationalists got all the reliability (i.e. A will never, can never equal not-A – it will always equal A).
In an attempt to resolve this schism, Immanuel Kant went further, identifying another similar distinction overlaid upon the old one; the analytic/synthetic distinction.
Analytic judgements are those where the predicates are entirely contained in the subject of the judgement, e.g. all killers have taken a life, all squares have four sides*.
Synthetic judgements are those where the predicates are distinct from the subject of the judgement, being linked through relation to another subject – e.g. the Earth orbits the sun, that piece of halibut looks good enough for Jehovah.
When overlaid, you end up with four categories of judgements.
- Analytic a priori – uncontroversial truths by definition – e.g. a triangle has three sides.
- Analytic a posteriori – a meaningless category because the predicates, being already contained in the subject, do not require empirical verification. E.g. to point at a real material triangle and claim “that triangle has three sides” is both rationally and empirically true – it necessarily has three sides by definition, and empirically it appears to have three sides. The rational truth is necessarily true, but the empirical truth, being empirical knowledge, is never utterly reliable and in this (and every other analytic a posteriori) case, redundant.
- Synthetic a priori (the point of this post) – judgements that are necessarily (a priori) true that have predicates derived through relations between subjects. I’ll get back to this.
- Synthetic a posteriori – everyday, empirical facts e.g. homoeopathy is just a placebo, thermonuclear devices work through a combination of fusion and fission, and Coldplay are boring.
Ignore for the purposes of this post, that my tutor at Uni who got me into Kant, had an example of a judgement that was both analytic and synthetic.
Why are synthetic a priori truths are necessary for the possibility of experience?
Here are some (modern) examples of what Kant would probably have considered synthetic a priori judgements.
- Non-Euclidean geometry.
- The evolutionary algorithm (if variant, if hereditary, if selected then evolution must occur).
- Graph theory.
None of the predicates within either of these reside just within the subject (e.g. hereditary status requires two subjects at least – a parent and a child), but all by definition are true. Their synthetic nature providing the practical justification previously only the province of empirical a posteriori judgements, while retaining the logical necessity afforded by their a priori status – a reliability previously only the province of Enlightenment rationalism.
All three of these synthetic a priori truths form the practical, logically consistent underpinnings of (scientific) experience; general relativity has non-Euclidean geometry as a logical underpinning; evolutionary biology has the evolutionary algorithm as a logical underpinning; a nice swath of computer science has graph theory as a logical underpinning.
Analytic a priori truths are too divorced from the practical, material reality to support any of these ventures. Synthetic a posteriori truths are the ventures themselves, not their logical underpinnings, and are insufficient in their logical necessity to reliably imply universality.
Analytic a posteriori judgements are meaningless, which leaves (in Kant’s scheme) synthetic a priori judgements – which as the examples demonstrate, do the job of logically underpinning experience just nicely (practical and by definition logically necessary) .
I hope that answers the question adequately. If all else fails, remember the four categories and the three that aren’t up to the task!
~ Bruce
UPDATE: Oh, and prior to Kant, where was also the fact-value distinction, which is outside the scope of the post – but worth knowing about.
* “God has existence” is not an example of this, because not only isn’t existence a predicate contained in the subject, it’s not a predicate at all. This is not a linguistic argument, it’s epistemological! Arrgh! (Pet peeve – sorry).
Oh, and on a related note – common wisdom and public mythologies about participants in the discussion of religion.
I’ve been enjoying a little trick of late, that I may formulate into a questionnaire of sorts.
Basically, I’ve been taking the work of religious writers, either stripping them of identifying details or presenting them to unfamiliar audiences, providing sources for the quotes they make and then prompting feedback. I’ve been doing much the same thing with chunks of text written by Dawkins, Hitchens et. al.
The people I’ve been presenting this material to have been those that identify as moderate-tolerant – either atheist or theist. Either shying away from ostentatious conversation, or subscribing to various tropes about vocal atheists being intolerant, purely engaging in ad-hominem, etc. etc. I’m an atheist, but… or I don’t subscribe to fundamentalism, but…
Interestingly, the Hitchens and Dawkins quotes have been well received (“I don’t mind this kind of atheism, but…”), while the religious quotes have largely been damned as fundamentalist.
The fact that the Hitchens and Dawkins quotes seem to be viewed as more reasonable than the work of Hitchens and Dawkins (principle of non-contradiction!), is only half as interesting as the way the religious texts (specifically, obvious misquotes of atheists) are being received. Specifically, they are being written off as fundamentalist, right-wing, “not like mainstream Christians”, or otherwise allegedly different to moderate Christianity.
So far nobody has sniffed at the end of one of these dismissals – although I live in hope.
The irony lies in the fact that I’m using religious texts by popular, well-respected, progressive, supposedly moderate, mainstream Christian authors! Alister McGrath, Madeleine Bunting and so on. I’ve even sometimes felt the need to defend these authors in the process when the condemnation has got a bit overzealous – “you know, the author actually isn’t right-wing”, or “no, no, they criticised the Pope for that!”
It’s all anecdotal of course. Not a controlled study. But it’s still interesting all the same.
I wonder what a proper study into perception-of-author-versus-perception-of-the-text would show! Maybe someone else has already done the research. Better get checking!
~ Bruce
I’ve got an email sitting in my in-box in response to my last post, and as it was so abrupt as to be largely devoid of tone (and from someone I don’t know), I can’t tell what spirit it’s written in. Hence, unless I’m given reason to think otherwise, I’ll assume it isn’t hate mail and therefore won’t publish the author’s name.
At any rate, their statement was in response to this line in my last post:
“I’m willing to help subvert the term until it’s rendered as meaningless and non-controversial as “afairyist”, or until I’m dead.”
(Me, Then.)
To which they asked “certain?” – and that’s all.
Before I get to the task of answering this question, I’m going to ramble on a bit about a discussion last night on Twitter. Like it, lump it or leave – you’re in charge of what you read.
Accusations of certitude
I suspect, but don’t know, that whoever our friend is, he’s probably been following last night’s Twitter conversation on the topic of atheists and certainty – where Christopher Hitchens and Richard Dawkins were nominated as examples of atheists exhibiting certitude. 140 chars isn’t much, and people were a bit tired so I didn’t labour the point that Hitchens is on record as stating that Victor Stenger is too confident in God’s non-existence (and even then I wouldn’t call Stenger certain*), and Hitchens has explicitly stated many times (and in The God Delusion) that his disbelief isn’t certain, what it would take for him to change his mind, and that certainty isn’t a good thing for scientists.
Certitude is an absolute. And while it isn’t easy to detect – because let’s face it, people can hide their certitude and passing yourself off as stubborn isn’t the best debating tactic – when certitude rears its head it’s really bloody obvious. The willingness to disregard evidence because it contradicts a presupposed, mystically obtained belief, is certitude. Claiming a 100% confidence interval with no room for human error, ever, is certitude.
Certitude isn’t a personality trait or a rhetorical disposition. It’s an epistemological facet. Being overbearing, loud mouthed and obnoxious, doesn’t make you certain. Being a bit of a curmudgeon doesn’t signify absolute confidence.
And that’s another thing – confidence isn’t certitude either unless it’s absolute. A hypothesis confirmed or denied at a 99% confidence level is pretty damn confident, but it isn’t certitude because by necessary implication, there is a small probability that the results are wrong.
So when someone is brash, pig-headed and confident, or you otherwise take exception to their personality, that’s no reason to accuse them of certainty. That’s not what certitude is.
If you think they are over confident, say so and then provide some evidence to back up your claim (or expect to be dismissed). If you find their personality obnoxious, which is a subjective thing, say so and give some examples (or expect to be dismissed). Even if you do present evidence and examples, expect them to be scrutinised. Especially when talking to someone who knows their shit!
The other thing that strikes me – relating specifically to the matter of the downgrading of evidence in lieu of mythology – is that in seeing these myths prosecuted in the past few weeks, I’ve repeatedly encountered the “accusation first – I’ll send you a link later”, phenomena. I know people are busy, and can be tired, and they can forget and so on and so forth. I’m not making an accusation here, much less about specific instances – it’s just that it’s happened a lot of late, which as a trend is a bit worrying.
Put more simply – I’ve been enquiring into people’s perceptions of atheists – not to recount in the book I’ve got planned**, but instead to test my own perceptions before going ahead with the project – and this is the list of the evidence and examples I’ve been given in the past few weeks.
- …
That’s. It.
In all instances, I had counter-evidence at the ready (I’ve got quite a long list of sources available at my desk here), but didn’t get to deploy them as discussion was in most cases side-tracked or shut-down for various reasons of convenience (I’m sure most were genuine, but I can’t honestly assert that all were – and I don’t know which were or weren’t). Whatever spirit these discussions were undertaken in, I don’t think anyone could rightly complain if I dismiss these assertions on the grounds of a lack of rigor.
It is an added irony, that the certitude trope is one that climate change denialists level against climate science – their “evidence” ranging from non-existent to ephemeral tropes about “The Green Religion”; their consideration of counter-evidence, such as statements of (non-100%) confidence by the IPCC, not undertaken. But none of the people levelling the certitude trope against atheists were climate change denialists as far as I could tell (and I did check and to some extent found the contrary – that they respected climate science).
None in as far as I have reason to suspect, would be duped by the allegation that the IPCC operates in certitude – which is smart and correct – but it renders their use of the trope elsewhere a double-standard. If, considering the evidence, these people wouldn’t be likely to call the IPCC “certain” – then why not consider the evidence that Hitchens and Dawkins aren’t certain and indeed, warn against certainty? (Or am I wrong in my estimation? Would these people be duped by the bare assertion that the IPCC is certain about matters of climate change? Would they fall for this denialist trope?)
Further to this, none of them are fans of Andrew Bolt – who could in as far as accusations and evidence are concerned, tell you a thing or two about the need for due care for the truth (i.e. rigor when making accusations). I’m quite confident about them not being impressed by Bolt’s defamation of Popovic, and a recognising that published allegations require evidence to accompany them. Even if you aren’t going to be sued, it’s the right thing to do.
I think in future I may add “#TweetLikeAndrewBolt” to re-tweets of these sort tweets in future***. If things don’t change.
Perceptions of personalities are readily formed by our own psychologies and I think we would all do well to realise that. Not only welcoming challenges to these perceptions from others, but actively challenging our own perceptions through the seeking and consideration of non-confirmatory evidence. Call it a safe-guard against prejudice. Or a bulwark against certainty about other people’s certainty.
Importantly I consider all of these people to be in general, reasonable people. But largely reasonable people can believe unreasonable things. Perhaps you thought I was talking about fundamentalists? I wasn’t.
Back to the quote
Was I exhibiting certainty in my statement “I’m willing to help subvert the term until it’s rendered as meaningless and non-controversial as “afairyist”, or until I’m dead.”?
It’s a comment about my current emotional position – “I’m willing”. It doesn’t prejudice a change of opinion down the road by saying something like “I’ll never be unwilling…”
I find it unlikely that I’ll be unwilling to change my stance on this matter until the stated conditions are met – that theists cut back their use of the term “atheist” as an invective, at least to roughly the same level as non-theists. (Some non-theists use “atheist” as a pejorative, it should be noted.)
Maybe my conditions will change over time – and although I can’t foresee myself becoming a theist, I suspect that I’d still support the Out Campaign with the “Scarlet A” even if I did “join the other side”.
This is confidence. This isn’t certitude.
I do wish an appreciation of this distinction were more widespread.
~ Bruce
* In The New Atheism, Stenger concedes that some God hypothesis are outside the realms of scientific analysis. Not that he’s stating that there are better ways to “know God”, just that there are things currently, and by definition perpetually, outside human reliable analysis - this is hardly a statement of certitude about God!
** If I were using any of these examples as content in a book, I’d have made a disclosure and presented a release form up-front. Not that I have to.
*** If you don’t find this even vaguely amusing, then I think you’re being too serious. Chill out. Take the comment in the spirit that it’s given. I’m kidding. I think. I’m not certain. Maybe I’m not. Kidding.












