Blog Coordinator

Knobe's X-Phi Page

X-Phi Grad Programs

« New Book from Tamler Sommers | Main | Levin on X-Phi »



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

Jorgen Hansen

This is very interesting.

I tend to think that phrases such as “Boeing expected to have to lay off workers,” may seem less *weird* to the folk due to their passivity when it comes to how they comprehend such phrases. (That is, most people seem to passively interpret such phrases.) For example, if a teacher asks a class of first grade students to choose between taking a longer recess or leaving school early, when five minutes later she asks, "Have WE made a decision, yet?" I don't think most people would think that this teacher is ascribing any type of mental state to the whole (over and above the parts) - the phrase has merely become another way of asking, "Did each of you agree about something?" And then we interpret it loosely or passively. If this is the case, however, then *why* we do this would also be interesting.

But it also seems to me that the folk have an extraordinary ability to ascribe certain states to groups. Racism, sexism, patriotism and many other "isms" seem to demonstrate this quite well. I wonder what the results would be if the subject (Boeing) were replaced with something like "America". And whether or not the phrase, "America felt anxious about going to war," would seem "weirder," given that we'd be attributing mental states not to a group of *Americans*, but to the country itself. Again, my impression is that the folk merely tends to passively interpret such phrases as "Many Americans are anxious about going to war."

jonathan weinberg

Neat study -- any hypotheses as to why there would be these differences between your results and Josh & Jesse's?

One worry: I'm not so sure about whether you can put that much interpretive weight on whether people went with singular or plural pronouns there. For example, it's my understanding that nouns like "team" are typically rendered as plural in the UK, and singular in the US, but I doubt that there are such profound differences in folk metaphysicses of mind across the Atlantic. Also, fwiw, I'm personally happy to treat the entities themselves, in such cases, as both distributed _and_ the proper subject of the propositional attitudes; a property P can apply to an entity E, even when E is constituted distributively, without P having to apply to the constituting parts of E instead of E itself.

Edouard Machery

Ingenious indeed. There are many explanations of the difference between your results and J&K's.

As Justin and I have suggested in our 2009 paper, the difference found by Jesse and Joshua has probably nothing to do with the ascription of phenomenal states vs. non-phenomenal states. When one looks at the sentences that are phenomenal (e.g., Acme Corp is getting depressed) and the sentences that are non-phenomenal in their first study (Acme Corp intends to release a new produce this January), it seems clear that the latter sentences are such that the reader can easily imagine a context that make them natural; indeed, a context is provided by the sentence! By contrast, it is difficult to imagine such a context for the former sentences; indeed, in contrast to the non-phenomenal sentences, no context is provided at all to explain the ascription of the mental states. This is sufficient to explain the tension between your results and Josh and Jesse's findings away.


Mark Phelan

Thanks for the encouraging comments!

Jonathan: Edouard’s (and Justin’s) suggestion could explain the difference between our and J&J’s results. Also, if people are treating group state attributions as a useful shorthand to subsume the mental states of individuals, the weirdness asymmetry may be due to familiarity with this shorthand as a way of attributing intentional—but not phenomenal—states to the members of groups. After all, the readers of the Wall Street Journal are much more interested in knowing the collective beliefs of Microsoft executives than their collective pains!

Regarding the point about differences of use between the UK and the US: This is a particularly helpful topic to raise since we’re thinking of doing some cross-cultural follow ups. To speak collectively, our thought was that, the important thing would not be how likely people in the UK (for example) were to use “they” for groups as opposed to Americans, rather, what’s important is how much more or less likely than Americans they are to use “they” for mental state attributions to groups as opposed to non-mental state attributions. That is, while some difference may be explained in terms of linguistic differences, it doesn’t seem that the tendency to treat attributions of mental states one way and attributions of non-mental states another is such a difference.

Finally, while your principle seems fine to me, what’s important here is whether or not people actually are attributing the property to the group as opposed to its parts. What we’re suggesting is that what was previously thought to be some evidence in favor of thinking that people were attributing intentional mental properties to groups may not be as good evidence as previously thought.

Jorgen: Right, we think what’s going on here is a similar sort of loose use as that you mention in your example. I think it’s pretty obvious why we do this sort of thing. It’s just a whole lot easier to ask, “Have we made a decision yet,” as opposed to asking, “Has each of you agreed with all the others concerning the recess/going home issue?”


Perhaps you should also look at the entity or group in question and see if this makes a difference. "Expecting" is something that a commercial company is known to do - it publishes business plans and forecasts and so on. "Feeling anxious" is not a formal activity we are used to it experiencing.

Whereas perhaps a different group—a film censor's board or a child protection charity—could more readily be accepted as "feeling anxious" about something.

Of the NSPCC (National Society for the Prevention of Cruelty to Children here in the UK):

"When the NSPCC conducted a study into child welfare in the UK, the NSPCC felt anxious about the living standards of children."

Phil Hand

I would agree with Matt's comment. Companies are often assumed to be coldly rational profit-seekers, which is why ascribing feelings to them seems odd. But in certain rhetorical contexts you might want to make the company seem otherwise: "Smith's Industrial Widgets feels the pain of this town"; "Walmart is thrilled that we've won the war, and so..."
I think these kinds of sentences would be recognised by most as deliberate rhetorical ploys.

Mark Phelan

Matt and Phil: I think I agree with you, but I'm not sure I understand the force of your posts. Do you mean to be offering this as a critique of our study, a defense of realism about group minds, or an explanation of Knobe and Prinz's original results?

Joshua Knobe

Nice experiment! This is definitely a very innovative methodology, and the results came out beautifully.

I wonder, though, whether there might be certain cases in which we do attribute a mental state to the group as a whole but in which we wouldn't attribute that state to any individual member.

For example, suppose that some members of the faculty want to hire Professor X but totally hate Professor Y. Then suppose other faculty members want to hire Professor Y but totally hate professor X. In the end, the department might compromise on Professor Z, whom no one loves but whom no one really hates either.

In such a case, I feel like it would be right to say:

The department as a whole prefers Professor Z to any other candidate.

But it seems like this attribution could not be understood distributively, since no individual member of the department holds the mental state in question.

Mark Phelan

Hey Josh, nice case! It makes me think of Levin's claim (in Josh May's post appearing just after mine) that: "the results of the most methodologically sound and philosophically relevant studies discussed in this volume could have been obtained from the armchair." I agree with your intuition that it would be okay in this case to say, "The department as a whole prefers Professor Z to any other candidate," and that this attribution could not be understood distributively. But I still think it may be that the attribution is anti-realist. I think that, in this case, its likely that the speaker is using the sentence as a stand-in for, "The department as a whole will vote out a job offer to Professor Z and not to any other candidate." In other words, a speaker of the relevant sentence would be taking the intentional stance towards the department. I think that this use of group-state talk is less common...that's why you have to set it up so carefully. Anyone just presented with this sentence and no background would be inclined to understand the attribution distributively.

We're looking into ways of testing whether people are sometimes taking the intentional stance when they use group mental state talk. I think your case suggests a useful methodology.


I think I mean it more as an explanation of the original results. I think your study is valid and useful, just perhaps not extensive enough to derive a strong claim from.

Here's what I think about realism about group minds: Most people are realist about it, in a fuzzy way. By which I mean: it is an everyday mode of seeing which is useful for us to employ. If we want to talk about proper ontology—the really real—then we have to differentiate frrom our everyday way of thinking and talking.

I think when you force a more detailed mode of enquiry upon people, make them think about the nature of the question with examples such as Josh's, many of them would realise that a group cannot really be said to expect or intend something, and might change their answer.

This is not surprising to me, because I believe it is no different to the way in which we build other artificial constructs around us for the benefit of everyday living. Compound objects (tables, chairs, cats) are in my opinion also just tidy shorthand for groups of things (simple particles) and have no metaphysical identity of their own (mereological nihilism).

We talk about these groups of matter as real objects and ascribe to them their own sorts of attributes - sturdiness, softness, cuteness - which, if we could see clearly, we would realise are just as nonsensical entities at the basic level of reality as Boeing's expectations or anxiety are.

I don't see the point of this study. Groups have intentional states insofar as the intention of a group can be expressed as if from a single perspective. Also, insofar as a group can have interests (which means, that a group can be wrong about its interests).

Since different sorts of groups are determined to have a single perspective in different ways, what it means for different sorts of groups to have intentional states simply means something different. For example, it doesn't mean the same thing to say that a democratically organized co-operative has an intention as it does to say that a hierarchically organized corporation has one - despite the fact that their intentions would be communicated in similar ways, and show up in the news media seemingly equivalent.

We can learn a lot about the "intentions of organization" by looking at the nature of organizations. We can learn very little about them by surveying people's intuitions - since peoples intuitions on the being of the unity of organization intention are highly manipulated by anti-democratic governmental and media pressures i.e. the war on union organization in America since Reagan.

The comments to this entry are closed.

3QD Prize 2012: Wesley Buckwalter