One argument made against the proposal to legalise assisted dying in the UK is that making this change might result in older citizens feeling pressured to choose death, increased pressure on people to think about and defend their existence, and the inevitable acceptance of voluntary and, then, involuntary and non-voluntary euthanasia. This kind of argument can be called a slippery slope argument. A
slippery slope argument claims that if we make a proposed policy change, other changes or outcomes will occur, and because these other outcomes are objectionable, we should not make the policy change. I am generally sceptical of slippery slope arguments and in this post I wish to register some issues with their use.
Some of my doubts about slippery slope arguments are based on them so often being false. Assisted dying and euthanasia legislation is a case in point. Probably the clearest data available on the likely effects of policy changes in this area come from the Netherlands, which had some non-legislative tolerance of these practices during the late twentieth-century and legalised them in 2002. Amongst the most recent publications show that euthanasia rates in 2010 were comparable to those in 2001 and 1995, the distribution across various important categories (sex, age, diagnosis) was stable throughout this period, and, perhaps more significantly, the number of end of life decisions made without the explicit request of the patient actually declined. There are different slippery slope worries to which this data does not speak and various questions can be raised about how to interpret the data and what it shows, of course. But it surely leans in the direction of rejecting the slippery slope argument.
There is also a more general difficulty to raise about slippery slope arguments. In essence, such arguments have the following structure:
- There is an argument for adopting policy X
- X will lead to Y
- Y is bad
- The bad of Y outweighs the good of X.
- Therefore, we should not adopt X
An important point to note about this argument structure is that it does nothing to dispute 1. It could be the case that the arguments made in favour of adopting policy X are poor arguments. But the slippery slope objection does not assert this claim. In this respect, it does not dispute that there is some reason to adopt X. In virtue of this, the slippery slope argument must be understood as posing a ‘countervailing consideration’ – a reason that is supposed to stand against and outweigh the arguments for adopting X.
Adopting this form, the slippery slope argument should be thought to start on the back foot. In footballing terms, the argument begins at 0-1 (losing by one goal) and an advocate’s aim is to turn the scoreline in their favour (2-1), or, at least, to level the game (1-1). Because it has this structure, it is advocates of the slippery slope argument who shoulder the burden of proving their position. This burden of proof pertains to each of the other three premises in the argument.
In relation to 2, advocates of the slippery slope must advance some plausible reason to believe that there is a causal connection between X and Y. Quite unlike how slippery slope arguments often proceed, it is inadequate for their advocates to conjecture bad consequences. The structure of their argument – by not challenging 1 – means that it is incumbent upon them to do the work of establishing 2. Most often, proving this claim will require empirical evidence of the connection between X and Y. Sometimes empirical evidence is hard to obtain. For example, it is sometimes argued that it is difficult to know what will happen if we legalise various currently illicit drugs. But even in such cases, there will be pertinent information and theoretical modelling that can be brought to bear on the subject and the onus is upon advocates of the slippery slope to bring it so to bear. At any rate, because those who are being challenged with the slippery slope argument otherwise have the upper-hand in the debate, they should feel no obligation to disprove the slippery slope argument unless it can be shown they have a serious case to answer.
In relation to 3, it is important that advocates of the slippery slope do not merely point rhetorically to possible Ys – alleging potentially connected states of affairs and presuming they are bad. They must provide reasons that these Ys are bad. An example of this issue is the euthanasia debate worry that people will feel pressured to think more seriously about the value of their life continuing. Although many people react to such a possibility negatively, there are others who think that it would be a valuable change. The same is true of the worry that legalising assisted dying would lead to the acceptance of voluntary euthanasia. While some people worry about the latter, others are positive about it. And, in virtue of the slippery slope argument shape, the onus of showing that the best arguments point towards the posited Ys being deemed bad, again, lies on the shoulders of the argument’s advocates.
Premise 4 in the above argument is important because, in order for X resulting in Y to give reason to reject X, it is not sufficient to show only premise 3 – that Y is bad. It must also be shown that Y is bad enough to outweigh the reason to adopt X. Such an argument will often not be straightforward. For example, if legalising euthanasia is valuable because it gives many people a desired freedom to choose how to end their lives, it may not be enough to highlight that it will also increase the pressure on some other people to end their lives. We could fully agree that such pressure is a very bad thing, but still retain that it is not enough to outweigh the good we would bring to so many other lives. Something similar to this point is often made about the grave tragedy of people being run over by ambulances that are, nevertheless, permitted to drive at high speeds for the great good this permission achieves. And, once again, due to the shape of the slippery slope argument, it is the advocates of this view who must do the work of showing that the slippery slope they caution is not such a case, but, rather, one in which the bads of Y outweigh the goods of X.
It is worth noting, moreover, that advocates of the slippery slope are not obliged merely to demonstrate the plausibility of one of these claims. In virtue of the argument’s shape, they must demonstrate the plausibility of all of them. If any of 2, 3, or 4 is not adequately defended, 5 does not follow.
To some extent, the suggestion that advocates of the slippery slope provide these defences amounts to nothing more than the straightforwardly reasonable demand that those advancing views provide support for their views. But the shape of the slippery slope argument – by not challenging premise 1 – makes this demand particularly pressing. Simply put: we ought to remember that the burden of proof for premises 2-4 lies with advocates of the slippery slope and that, until they take up this challenge, the score remains 1-0 to their opponents.
Very nicely put. I have one small comment, and that relates to the level of proof of plausibility that one demands for the parts 2-4. It seems reasonable that these levels of proof need not be independent: if one has solid proofs of 3 and 4, for example, then perhaps slightly less solid proof of 2 ought still to be considered under certain circumstances.
For example, if one has established solidly that Y would be utterly catastrophic compared to the good of X, then perhaps even quite tenuous proof of the plausibility of 2, that X will lead to Y, ought to be considered seriously. (To put some numerics on it for illustrative purposes, if one only feels 10% ‘confident’ in X leading to Y, but the cost of Y is huge compared to the benefit of X, then it is reasonable to consider the possibility further.)
Hi Olof, thanks for this comment. I agree with you on this point. I guess we would not want to push it too far. For example, there are sometimes claims that policy changes in areas like euthanasia, abortion, same-sex marriage will lead to extreme changes in climate or generally apocalyptic scenarios. These Ys are clearly on the very bad end of the scale and it is the (complete) weakness of the link to X where the argument fails, and, therefore, where we should insist users of such arguments are obliged to do the work if they wish to be taken seriously. But I take it that you had in mind only somewhat relaxing standards of proof on some premises if other premises are clearly proven – something like an interdependent scale where the average must remain above a certain level. In this case, I think I agree with you. The nod on this point is much appreciated.
I share your skepticism about slippery slope arguments in general (though of course some have merit, the general argument you advance doesn’t undermine specific instances where all the premises are plausible) but I think this general formulation misses an important kind of slippery slope argument.
While your premise no. 2 is interpreted in empirical terms, I find that it’s often a moral kind of argument. X will lead to Y not as a prediction but if we cannot defend X, we have no moral grounds to reject Y. This is the kind of argument that people in the US (include former supreme court justice Scalia) have made regarding same-sex marriage: if we cannot prohibit that, we have no grounds to prohibit pedohilia, polygamy or incest.
Now, much of your argument also holds regards those kind of cases. Is incest really that bad? We have to think about that. Same sex marriage was the specter that Scalia held over his audience’s head in 2003 when they struck down laws that prohibit sodomy, and as a society we don’t generally think that’s so bad anymore. I know of some people who think that the moral aversion to incesnt is exaggerated (regardless of how it makes you feel) and there are definitely voices calling for legalizing, or at least accepting socially and morally, polyamory and multiple marriages (though that is distinct from polygamy as it has been traditionally practiced and thankfully I don’t know of anyone defending pedophilia).
However, there are other ways in which these kinds of arguments are distinct. They don’t require empirical evidence, but a normative investigation. They are about what the ways different moral stances are related. I’m of the opinion that many of these arguments fail partly because I believe hypocrisy is over-rated, but arguing against them is much harder than the way you depict. Rejecting the slippery slope often depends on the ability to articulate a position that justifies X and rejects Y, unless one is willing to accept Y. In many cases, as I have noted, you may have no problem with Y in principle but you know your opponents reject Y vehemently, and since X is very important – the political struggle for X might require figuring out a creative way to argue for X without also arguing for Y. In my experience, that’s a common feature of the slippery slope in political life – gay marriage, gun safety, trade deals – in all these cases the argument is often, ‘if we agree to X, we have no grounds to rejecting Y’.
Tomer, I had a similar idea when I came across the term slippery slope. I’m sure I’ve heard the term applied to the _reasoning_ the led to X, rather than to X itself. That is, “that’s a bit of a slippery slope — if we allow the argument that leads to X, then we also have to allow Y as it also follows from that argument, just applied to a different starting point.” I took this to be distinct from what the post considers to be a slippery slope argument — in particular, above one *does* question point 1 in the post, not through direct criticism of the argument’s individual steps but through pointing to (hopefully) absurd consequences. But perhaps this is not accurately labelled by the term ‘slippery slope’ — up to the original poster to say.
Whether it can be labelled as slippery slope or not, what you say in your final paragraph still applies: as a proposer of the argument for X, if you want to avoid these kinds of objections then you really want to isolate the parts that lead to only X or to as small a net surrounding X as possible.
Thanks for this comment, Tomer. To some extent, I did want to set aside these kinds of ‘philosophical slippery slope’ arguments here. It is difficult to address both these and the more empirical slippery slopes I had in mind in one post, and because the latter are also quite widespread, particularly in debates with which I am currently engaging, I took them as my main target.
Nevertheless, as you say, I think much the same argument format can be used to address these cases too, most directly by a modification of premise 2 along the lines that “the reasoning leading to allowing X also leads to allowing Y”. I can think of plenty examples that use this kind of reasoning that are similarly poor in demonstrating the case. For example, it is sometimes claimed that the reasoning leading to permitting voluntary euthanasia commits us to permitting involuntary euthanasia. Given the obvious normative different between these practices (to wit, voluntariness), which often features in defences of euthanasia, I would retain that it rests on the shoulders of slippery slope advocates to make these cases before advocates of the policy change have a case to answer. You are right that it changes the argument from an empirical to a normative one, but I think my thread about where the burden of proof lies in the argument remains accurate.
Fight for Freedom. Stand with Hong Kong
Thank you for the sensible critique. Me & my neighbor were just preparing to do some research on this.
We got a grab a book from our local library but I think I learned more clear
from this post. I am very glad to see such wonderful info being shared freely out there.