On Justification and Argumentation
In a previous post, I wrote about abortion, which of course received some amount of pushback. However, I also got pushback from an unexpected direction. A number of commenters pushed back against trying to get more complete justifications for their beliefs.
Bsycho for instance says:
Does it harm the life or liberty of the uninvolved party?
If the answer is “no”, then the uninvolved party can fish off. Period
And when pushed as to why this is the case, answers:
Because harm to them is the only legitimate criteria by which they have any standing whatsoever to care. Without harm they’re just throwing a hissy fit because the world isn’t conforming to their personal aesthetic tastes. A society attempting to accommodate such people is flushing freedom down the toilet.
I don’t see how Rawls changes anything here. If you’re A, you have the right to do whatever you wish with consenting adults; if you’re B, you have that same right. If you’re B and you whine because you don’t get to block A, then you’re confusing liberty with obedience.
Charitably interpreted, I think that bsycho believes that his answer is sufficient to answer my question. Unfortunately, I disagree with him. One minor quibble is about whether the characterisation he provides is true. After all, we can cash out both sides in terms of liberty. One can on the one hand be free from being enslaved, or on the other hand be free to keep slaves. In cases of conflicting freedoms like this, it is important to explain why the freedom from enslavement trumps freedom to keep slaves. However, the key point of disagreement here is that the argument he gave is not sufficient. I can still ask why we should care about liberty instead of obedience. i.e. there is a problem of how to stop the sceptical regress.
I’ll try here to show why not being able to stop the sceptical regress can be problematic. When we do these kinds of things e.g. saying that we should legalise X or have this kind of policy and someone disagrees, presumably, the ensuing conversation is aimed at finding out who is right or wrong, or maybe showing why you are right and the other person is wrong (if there is a distinction to be made between the two ends). Productive conversations therefore contain arguments which feature reasons. Insofar as the aim is merely to convince, you start out from premises your interlocutors will agree with and from there proceed to make your argument. After all, if your interlocutors cannot even agree with your initial premises, then even if the logic of your argument is valid, how can you expect to convince them of said conclusions?
That is not all though. Just because you and your interlocutors agree on the premises, it does not follow that your premises are true. There is a certain futile aspect to arguing for a position without caring whether or not your premises are true. (What is the point of trying to convince someone if you do not think your position is true?) In certain ways then,
Nevertheless, people still seem to disagree even when they share the same premises. This disagreement is due to them drawing different conclusions from the same premises. The problem with this is that if two conclusions can be equally reasonably drawn from the same premises, then the premises in under-determine the conclusion in such a way that neither of the conclusions in question can be said to be drawn from the premises. (Though a weaker more general conclusion which is consistent with both of the former conclusions may still be said to be drawn)
For example, two people may differ on what the results of choice in the original position are in that one person says endorses Rawls’s 2 principles of justice in their entirety while another believes that something other than the difference principle would be chosen (e.g. Twin Rawls vs Real Rawls). However, even if we consider that both are equally and maximally reasonable, it does not follow that no conclusions may be drawn from the original position, only that any conclusions that could be reasonably drawn would have to be so general as to accommodate both views. (That is assuming that both are supposedly equally reasonable views) This means that in order to argue for one conclusion over another, you must show that the connection between your conclusion and the premises are stronger than that between your interlocutor’s conclusions and the shared premises. Judgements about these can get complicated, but we can make a few simplifications. Whatever else may be true, deductive arguments provide the strongest connection between premises and conclusions. The conclusion of a valid deductive argument necessarily follows from its premises.
Inductive arguments are more complicated, but statistical theory gives us a good handle on exactly what conclusions we may draw from our data. However, given that we’re talking about political philosophy (What should policies aim at, what is the role of government etc), induction is rarely if ever employed. In this blog format, we at best quote the empirical work done by others, we don’t process policy results ourselves.
Abductive arguments or arguments to the best explanation are even more complicated than induction. B explains A iff p(A|B) > p(A) where A is a given and p(A) is the antecedent probability of A. The problem, however, is that what counts as a better explanation is not always so straightforward. Ideally, we would like p(A|B) = 1 and B to be independently likely to be true. However, this often trivially obtains when B=A. For example if A is the observation that the grass is wet, then ay explanation for why the grass is wet would be such that grass would be more likely to be wet if the explanans B obtained than otherwise. For example, raining the previous night would make it more probable that grass was wet. Someone sneaking into our house to water the garden would also make it more probable that the grass was wet. Now, these things could only be explanations if the grass was relatively unlikely to be wet otherwise. However, the latter possibility (the sneaky gardener) is less likely than the former and therefore makes for a worse explanation. Of course, very trivially, the statement that the grass is wet because there is water on it satisfies the formal conditions eve better, but is really just a restatement of the explanandum A. In this context, Stillwater’s comment is apposite:
But at some point the argument’s I make or the views I hold don’t derive from initial premises. They devolve to adhoc-ery. Which isn’t to say that ad hocness is an inconsistency. But when you include too much adhoc-ery, one’s theory gets very close to simply a complete description of preferences. And that’s not a theory anyone could be proud of.
Another deeper problem with abduction when it comes to this kind of normative reasoning is that the explanation so to speak is being used to justify the very “facts” that we used to justify the explanation. i.e. if I am trying to determine whether or not the grass is wet outside, it doesn’t make much sense to try to work backwards and ask what would be the best explanation for the grass being wet. Rather, we could rather look outside and if that is not possible, try to determine if any of the possible causes of wet grass actually obtained. But that is an entirely different kind of thing from trying to find the best explanation for a given phenomenon. We can see how this would go wrong when we fill in more details. There is a whole list of positions out there: i.e. on abortion, trade, regulations, unions, the death penalty, abortion, gay marriage etc.
As Stillwater intimated above, any perceived inconsistency can be trivially resolved by invoking a principle which would distinguish between the situations. However, very quickly, the invoking of such principles becomes extremely ad-hoc such that a complete statement of one’s principles just is a statement of one’s political preferences. The problem with having ad-hoc principles is that there is basically nothing else that justifies such a principle (and the conclusions themselves cannot justify the principles). Then, in that case we have no reason to hold on to said principle and by extension, the conclusions that the principle “explains”. While such a method may give us the most plausible principle that justifies our beliefs about justice and morality, it does not give us any way to adjudicate between different moral beliefs.
Now, of course one possible objection to this criticism of abduction is that our initial moral judgments are good enough to serve at least as a starting point. Principles are chosen in light of these judgments and may if sufficiently well supported by the majority of our strongest judgments, be used to revise some other judgments. In this view, when moral theorizing, a moral/political theory is better if it keeps to most of our common sense judgments even at the cost of some contradictions or ad-hoc features. i.e. here, taking things to their logical conclusion is something we shouldn’t do.
It is in light of the above kinds of objections that the following kinds of comments make sense.
I would guess that most people don’t start from first principle reasoning of what an ideology should be and construct their whole views, opinions and belief system based on that. Most people believe certain things about certain issues, and then look for which political ideology or political parties best fit their views on those issues, making compromises along the way
I scorned socialists for that very reason- that they let an abstract ideology define their beliefs, and followed it to its conclusion regardless of how absurd the outcome
And in a later post, Tod Kelly agrees:
Take almost any position and if you follow it to an extreme, it’s often best to be able to say “yeah, well not that.” Refusing to budge on any little thing for any little reason in the name of purity of thought is the stuff of dystopias
As do Jason Kuznicki:
I actually agree, mostly. I’m a Hayekian, not a Rothbardian. I don’t do first principles and deduction. I much prefer lived experience and incremental change
And Pierre Corneille
I do think it is possible to be so caught up in consistency that one comes off as an absolutist, uncompromising person. I have a trotskyist friend is who is very consistent in his views (with some exceptions, as I imagine is true of everyone), and he’d be quite a scary dude if his proposed revolution ever really had a chance to take off.
The thing is I don’t think our moral intuitions are worth jack shit. People differ vis-à-vis their intuitions so much on so many issues that they usually contain more heat than light. And it is absolutely mysterious how people expect their gut to have any access to moral truth? Why is it that when your arguments start leading you to “absurd” results it is somehow indicative of something having gone wrong? I hate to sound like an objectivist, but relying on your commonsense judgments is just anti-reason. There is no rational basis for doing so. In order to make any kind of incremental change, we can only know that we are taking a step, however small, in the right direction. However, there is no way to know whether the direction you are going in is the right one without working out in advance what your views are. Even more importantly, in order to take the view that we should be incrementalists rather than radicals, it is not enough to say that radicalism produces certain consequences. Nor is it enough to say that many people think that these consequences are bad. We must be able to show that they are bad, or unjust or that the founding fathers thought it was bad. i.e. when someone comes up to you and says: We find these principles to be self-evident… you don’t just uncritically accept said principles, you ask whether they really are self-evident. Yes I am the guy who will probably respond that I don’t find them self-evident. That people are endowed with certain rights in virtue of being persons may be a true proposition, but it is not self evident. And it is a failure of reason to fail to question whether such pronouncements are in fact true.
None of this is to say that I am immune to error, only that if we do find ourselves facing a contradiction, or holding an ad-hoc principle, we know that we are wrong or fail to adequately justify our beliefs. What this means is that accusations of inconsistency or ad-hocness are serious accusations. A valid argument proceeding from really self-evident premises will contain true conclusionsand be indefeasible in a certain kind of way that cannot be defeated by thought experiments, trolley cases etc. Whatever rightful purchace such tools have on our reason, they cannot be more than the result of a sound deductive argument which is necessarily true.
This brings me to address one last point by Mr Kelly:
Yep. This is pretty much Philosophy in a nutshell. Rewriting reality to make it fit more neatly into your clever world view.
What, liberals aren’t for all those thing? No, no, don’t admit error. Just go back and make your argument longer, and use bigger words. You’ll get there!
The difference I have with Tod is that I think that it is insufficient to take liberalism to be what self-identified liberals believe. There are two reasons. The first is that most people lack consistent beliefs and taking what they believe to be liberalism as liberalism is kind of like using a strawman as I could easily use the nearest consistent version. The second reason is the Americans use the term liberalism idiosyncratically. Liberalism has become a smear word for the entire left. I wouldn’t necessarily characterise Mr Kelly, Stillwater et al as liberals. Rather, they are some kind of moderate egalitarian pragmatist. The care a lot about equality and the common good and they search for policies that best realise their goals of increasing equality and the common good. Now, I am not arguing that this is wrong (but it has to be because I am not an egalitarian pragmatist! *grin*) All I am saying is that the sum total of their political beliefs is better characterised by something else other than liberalism, which has a very specific meaning, and which refers to a group of ideologies with certain central concerns which are themselves somewhat contested. Now, of course I will have to make an argument for that which I am not making here. What it boils down to is this: I will say that ultimately X or Y or Z are fundamental liberal values because embracing them gives us policy positions that cohere with most of what we take to be provisionally liberal views. If we really cared about those values, we would embrace W, which we don’t. Now, my argument for W may ultimately not be successful. For example I suspect that that could be possible about my abortion argument. But because the argument uses concerns that are central to liberalism to make the move it does, whether or not it is successful, it is a liberal argument.