Blog Coordinator

Knobe's X-Phi Page

X-Phi Grad Programs

« Actually, women are MORE likely to share their (mostly-male) professors' intuitions... | Main | Call for Unpublished Studies on Means/Byproduct Distinction »



Feed You can follow this conversation by subscribing to the comment feed for this post.


Thank you for this paper. It is sad the lengths the field will sometimes go, including seemingly never ending games with “intuition” to avoid engaging with incredibly obvious attempts to improve inquiry.

Joshua Knobe

Hi Jenny!

So nice to see this paper. It's beautifully well-argued, and it's wonderful to see it coming out in Thought. Congrats!

Anyway, I guess I have a slightly different view about the present dialectic surrounding these questions. In your paper, you take up objections to the negative program that seem broadly metaphilosophical in character (e.g., questions about the epistemological significance of certain findings). However, I am thinking that these are not even close to being the most important objections in the existing literature. Rather, the most important objections right now are clearly the *empirical* ones.

Early empirical work on these topics suggested (a) that people's philosophical intuitions were predicted to a shocking degree by demographic characteristics (ethnicity, gender, etc.) and (b) that people's philosophical intuitions could be manipulated by incredibly subtle contextual cues (e.g., priming disgust to change moral judgment). These results suggested that people's philosophical intuitions were far less reliable than we had previously thought, and this prompted some complex and difficult metaphilosophical questions.

But subsequent empirical work seems to be going in exactly the opposite direction. The majority of the early studies alleging demographic effects have failed to replicate, and there have been a bunch of surprising studies pointing to ways in which people's intuitions are strikingly robust across demographic variables. Similarly, there has been a real revolution in social psychology, with researchers recognizing that the studies that had originally been taken to indicate effects of subtle contextual cues rather consistently fail to replicate and seem to have been flawed in certain fundamental respects. Overall, one might reasonably conclude that the totality of findings point to the conclusion that people's intuitions are actually somewhat more reliable than we would have expected prior to the time we started doing any studies on the matter.

At least on the surface, it seems then that this more recent research is pushing us toward a completely different set of metaphilosophical questions. For example, the amazing recent paper by Machery et al. seems to prompt the metaphilosophical question: 'What are the metaphilosophical implications of the fact that epistemic intuitions are so shockingly *similar* across different cultures?' In much the same way, the recent Landy & Goodwin paper prompts the question: 'What are the metaphilosophical implications of the fact that people's intuitions are so strikingly *unaffected* by contextual cues that trigger disgust?'

Of course, I am super open to the thought that this initial impression might be incorrect. As alway, there is plenty of room for surprising objections which overturn what initially seems right, and I would be really interested in hearing from researchers who think that, despite this surge of recent studies, there actually is good reason to keep pursuing metaphilosophical questions about the implications of unreliability in intuition. All I mean is that it seems clear that this empirical sort of objection should be the first battleground for anyone trying to advance something along the lines of the original negative program. Before we can engage in a serious way with questions about the implications of certain effects, it seems like we need to establish the existence of those effects themselves.

In any case, thanks again for a really nice paper!

Jennifer Nado

Thanks Josh! I agree that the empirical challenges you mention are much more serious. But one thing that I'd suggest is that changing the way we talk about these issues is worthwhile even here. For instance, I don't think we should take the studies we have to indicate e.g. that 'intuitions are stable cross culturally' or that 'intuitions are unaffected by disgust cues'. Different intuitions will be different. Likely some will be stable, others not. Some will be crossculturally varible, others not. Even within a category like 'epistemic intuition', I doubt we have enough data to draw a generalization like the one you suggest above. Maybe intuitions about the epistemic import of reliability e.g. work differently from intuitions about whether knowledge is the norm of assertion. I'm opposed to drawing broad conclusions from the data on either side. The full picture is likely to be tremendously more complex.

Joshua Knobe

Hi Jenny,

This response is super helpful. Basically, it seems like we can take the framework you developed to understand what people used to think was the empirical finding and then just turn it around so that it can be used to understand what now it appears to be the actual finding.

For example, suppose someone says: 'Existing studies show that people's epistemic intuitions are shockingly robust across different demographic groups, so we have to conclude that people's intuitions are more reliable than they appeared to be prior to these experiments.'

You could respond: 'Wait! It is true that existing studies find remarkable similarities in epistemic intuitions across different ethnic and gender groups, but it is a mistake to think that there is some one single psychological process underlying all epistemic intuitions. So even if we find no demographic effects for these particular epistemic intuitions, we might well find one for other epistemic intuitions.'

Jennifer Nado

Yep, that's the idea. Obviously there's a worry that that kind of move could be abused - we do want to be able to make SOME generalizations beyond immediate results of studies, after all. But my thought is that, at current, we don't know nearly enough about the psychology underlying these judgments to fix on the relevant 'categories'. Ideally, we want to be able to infer from a given study that similar effects are likely to be found for other judgments produced by 'similar enough' psychological processes. But at the moment I don't think we have enough understanding of the various sorts of intuitive cognition to say which psychological processes are 'similar enough', for basically any category of intuitive judgment.

Wesley Buckwalter

Despite some high profile replication failures, see Edouard Machery's forthcoming book for an up to date literature review of quite robust demographic and presentation effects relevant to these matters. I don't think it's accurate to summarize this literature as "going in exactly the opposite direction", but perhaps Jenny's call for more specificity is exactly what will be helpful in deciphering these findings and moving forward.

Joshua Knobe

Hi Wesley,

Thanks so much for this comment and also for everything you've done in your own work to advance our understanding of these issues.

I'm not sure if there is anything we really disagree about here, but to begin with, I just wanted to note that we probably agree about two things.

1. There are specific cases in which experimental philosophers have found demographic effects. For example, you yourself found a gender difference in intuitions about brain in vat skepticism.

2. It would not be accurate to describe work in the experimental philosophy on these issues quite generally by saying that it supports the existence of demographic effects. Over the past five years or so, there have been a whole bunch of studies with findings that specifically go in the opposite direction, providing evidence of surprising ways in which people's intuitions are *not* impacted by demographic factors .

Taking all of the evidence together, my sense is that existing results suggest the following conclusion: Overall, people's philosophical intuitions are less affected by demographic factors than we would have expected them to be prior to the advent of experimental philosophy. It really is quite surprising how many of the effects uncovered in recent research arise in people from different cultures, in people speaking different languages, even in very young children. Just to give one example, I find it extremely surprising that the epistemic side-effect effect (ESEE), which you discovered a number of years ago, also arises when the study is run in China or South Korea.

At this point, then, I feel like we face questions on two levels. On the one hand, we face questions about the individual cases in which there are demographic effects. For example, "What are the philosophical implications of the fact that there are gender differences in intuition about brain in vat skepticism?" On the other, we face questions about the broader pattern. For example, "What are the philosophical implications of the fact that the patterns in people's intuitions about knowledge are generally so surprisingly robust across cultural and gender differences?"

Of course, I would be very open to arguments against this view. A careful look at the empirical findings might show that initial impressions are mistaken and that there is actually some arguments against, e.g., the claim that people's intuitions about knowledge are surprisingly robust. I would be very interested in hearing such arguments. What I object to, however, is the practice -- found only in people who are not themselves experimental philosophers -- of just ignoring these empirical issues entirely and jumping right to the part where one discusses the meta-philosophical implications of surprising cross-cultural difference.

The comments to this entry are closed.

3QD Prize 2012: Wesley Buckwalter