The Collective Conscious

Consider the sentences:

1. Dmitri is experiencing great joy at his new job.

2. Sony is experiencing great joy at its increased sales.

If you're like me, your immediate intuition is that the first of these sentences makes a lot of sense while the second is completely ridiculous.  A corporation like Sony can't be experiencing great joy! Group agents like corporations just aren't the sort of entity that are capable of phenomenal consciousness.

In an exciting new paper, the philosophers Bryce Huebner, Michael Bruno and Hagop Sarkissian show that this sort of intuition is not nearly as universal as one might have supposed.  In fact, they show a dramatic cultural difference between the intuitions of Americans and the intuitions of Hong Kong residents.  Americans think the first sentence sounds a lot more natural than the second while Hong Kong residents think that the two sentences are fairly similar in the degree of naturalness.  (Even Hong Kong residents think that the first sentence is better than the second, but this effect is not nearly as strong in the Hong Kong sample as it is in the American sample.)

Drawing on these new results, Huebner and colleagues argue that there are important cultural differences in the way people think about individuals versus groups.  Americans see a yawning gulf between individuals and groups -- individuals can be conscious and groups cannot -- while Hong Kong residents think that group agents are actually fairly similar to individual people. 

Comments

[Ryan]

I would like to ask a question that is slightly broader than this post. What is the positive program of this sort of experimental philosophy? I think that it is important that previous work has uncovered that relying on intuitions as solid evidence for philosophical work is a bad idea, as intuitions vary across cultures and demographics. I am sure that there are many particular instances of this broader fact that we have yet to uncover. This most definitely should shape how philosophers go about their work. But so far this is just a negative claim - what is the positive program?

It seems like surveying "the folk" to find out how they use terms is not going to be terribly productive. For one, they can only tell us what people do think, not what they ought to think. While this may be important in some areas of theory construction (for example, what people think normative is good to know), it seems unhelpful in other areas (it doesn't seem to matter whether the folk thinks that machines can have consciousness, as it's not getting at the issue of whether machines could or not).

I think this matters because it seems that experimental philosophy is not being suitably different. Finding out the usage of terms is just a replacement for concept analysis in traditional analytic philosophy. But it seems that this is a wrong model for moving forward. Replacing our intuitions with the intuitions of the folk doesn't seem to be an improvement. Instead, shouldn't "experimental philosophy" be, well, more experimental?

[jonathan weinberg]

Amen. :-)

[whitebeard]

"For one, they can only tell us what people do think, not what they ought to think."

"...tell us...what they ought to think"

an ethicist would tell you that "oughts" apply to behavior not to thoughts. the proposition that someone/anyone could tell someone else what they *OUGHT* to think is bizarre, if not frightening.

[whitebeard]

"Sony is experiencing great joy at its increased sales."

It seems to me that this sort of statement increases the likelihood that the recipient of it will not correctly associate the feelings/behavior to responsible human beings, but rather to an abstract object which cannot be (easily?) held accountable.

[alexandra]

Joshua,
this is a very interesting and exciting result! I'm looking forward to reading the paper. Off the top of my head, I'm wondering what kind of experiments have been done on any cross-cultural differences in attributing intentions to collective vs individual agents. Anecdotally, it seems that Americans are pretty willing to attribute both intentional actions and moral responsibility to collectives-- Enron, American Airlines, etc. But as a former professor of mine loved to remind me, 'anecdote ain't evidence'. So is phenomenal consciousness unique in this respect? I guess I should just read the paper and find out...

[mhsu]

Does this have more to do with the awkward phrasing of the sentence? If you change it from

"Sony is experiencing great joy at its increased sales."
to
"Sony ecstatic about increased sales."

It would look like a fairly standard headline in any business section of a newspaper. From my experience students or the general population in HK are usually not so sophisticated at English to tell the rather subtle difference between the two.

More generally, there are quite a few papers on the proclivity of investors of all background to "personify" the market. Now there might well be some cultural differences, but it certainly isn't the case that "Americans see a yawning gulf between individuals and groups -- individuals can be conscious and groups cannot". Anyone who has followed the Bear Stearns meltdown will endured the great number of puns on the various ways that a "Bear" can suffer.

[Edouard Machery]

Great work!

It would have been nice, though, to acknowledge the (in my mind pretty serious) objections against Josh and Jesse's study by Adam Arico and by Justin Sytsma. Particularly, since you are making exactly the same assumptions and asking the very same questions, you face exactly the same problems.

Edouard

[Bryce Huebner]

Josh: Thanks for posting this stuff. I am swamped right now so hadn't taken the time to put it up.

whitebeard: You are right that participants were supposed to associate the feelings and intentions with the groups and with the participants; we worked hard to word our prompts in such a way that the distributive readings of the ascriptions were made as problematic as possible.

Alexandra: There is indeed a ton of work on cross-cultural differences in the ascription of intentional states to collective and individual agents. We go through some of this evidence in the paper and you'll find some citations in the bibliography. Basically, the finding is that while Americans TEND to think that individuals are the only sorts of agents that can cause social outcomes, East Asians TEND to thing that groups can be the agents of social outcomes as well.

mhsu: You are dead on with the stuff about the personification of the market; and, I think that there is a rich philosophical question about whether collective consciousness is ruled out by a more Western understanding of the world. Indeed a chapter of my Ph.D. thesis was dedicated to answering precisely this question. In this paper, however, our goal was to reply to the paper by Joshua Knobe and Jesse Prinz in which they report data that show that American subjects were willing to ascribe intentional, but not phenomenal states to groups. Perhaps you are right that the sorts of questions we asked might be doing some of the work in the difference here. However, our questions was whether participants would be willing to ascribe *the same sorts* of mental states to individuals and groups. Given that "Bob ecstatic about product sales" sounds a bit weird, we just went for something that sounded more straightforwardly like a run of the mill ascription of a mental state.

Edouard: You are right that we need to work through the objections from both Justin and yourself and from Adam. I intend to add this stuff back in when we revise the paper. As it sits, this paper was written quite a while ago (read: before I had seen any of this stuff), and we haven't had the time to talk this stuff through as a group and to get back to the paper to add this stuff in.

Thanks for the comments and I (and I would assume we) look forward to hearing more!

[Ryan]

whitebeard: I'm not sure I follow you. Doesn't some knowledge of algebra tell you that you *ought* to think that 1 + 1 = 2? Or some familiarity with biological and archaeological evidence suggest that you *ought* to believe in evolution? Large chunks of epistemology try to establish precisely what the conditions are for when you ought to think one thing and not another. Epistemic rationality places some conditions on what one ought to think, if one wants to keep up good inferential practices. Oughts of course apply to beliefs and thoughts, not just to actions. It isn't to say that there aren't also matters of taste, etc, but I take it that the enterprise of philosophy depends on arguments, and arguments are supposed to show the reader that they ought to think the conclusion given the premises. I'm not sure what's either bizarre or frightening about that.

[arlo]

and yet, sony shares a lot of the same legal rights as dmitri. funny how the law doesn't draw a distinction, but the average american brain can.

[Josh Weisberg]

Interesting stuff.

Bryce:

1. Did you control for language differences? Did all the Hong Kong students speak English as a first language? Did the American students? That's a source of potential interference, perhaps. If non-native speakers are translating into another languages with different terms or distinctions, that could pose some problems.

2. Just curious: why did you have pride as non-phenomenal on your questionaire? Seems to have a distinct phenomenal component to me, as far as that goes.

3. Relatedly, I wonder if you'd get Knobe type effects if you ask Americans whether Germany feels guilt, or if Iran experiences great glee at the America's troubles in Iraq, or something like that. I wonder if moral attribution of responsibility for something bad might make folks attribute something like guilt (if that's phenomenal) to groups. Likewise for anger, jealousy, glee (kind of a gloating happiness) etc.

4. I guess I share the worry that the attribution of feelings to groups may be hidden attribution of feelings to the individuals that make up the group. And I'm not sure how to pull that apart methodologically. Could SONY experience great joy if none of it's employees or owners do?

Anyway, great stuff.

Thanks

Josh

[Bryce Huebner]

Josh,

1. You are indeed right to worry about the language differences issue. We did not control for this in any rigorous manner. As far as I remember, the participants were not native English speakers (Hagop would be the one to ask here as he is the one that collected our data in Hong Kong). While you are indeed right that this is a "source of potential interference" I, at least, am inclined to hope that it's nothing more than potential threat.

I was actually inclined to read the fact that we got these results from subjects in Hong Kong as pretty strong confirmation of the result as these are likely to be the most Western of any sort of East Asian participants we could have used.

Hagop, I would love to hear what you have to say about this if you want to chime in though!

2. I think that you're right that pride is a funny case. our inclination at the time was to think that 'being proud' didn't require anything more than the intentional state while 'feeling pride' seemed to get the phenomenal component back in. I'm not sure that i would still think that now, given the results that I collected on ascriptions of mental states to robots and given the stuff that Justin and Edouard have been working on But that was a working hypothesis given the data that Josh and Jesse had collected.

3. I really like your suggestion of looking at cases where people are willing to make the moral judgment about the corporate agent. I am inclined to think that something like moral judgment might just get people to ascribe the purportedly phenomenal states to groups. It's well worth running and I'd love to try it out. we should chat more about this though--drop me an email!

4. I've also got some thoughts about this stuff on trying to pull apart genuine ascriptions of feelings to groups and ascriptions of feelings to the individuals that compose the group. Philosophically speaking, I've got arguments for the claim that Sony can experience joy even if none of its employees do? However, in order to get to the commonsense ideas, I think we have to be pretty creative. When we gave this paper at the SPP last year, Chase Wren suggested that we try to look at the acceptability of ascriptions in cases like this: "Bob is feeling angry and so is (Sally or Sony)" or "Microsoft wants to increase profits this quarter and so does (Apple or Agnes)".

I think that it's a good idea to run some of these, but I haven't sat down to design the experiment. Perhaps I'll see if i can get Chase on board and try to run this up here in Boston and down there in Alabama...

By the way, thanks a bunch to the X-Phi community for all of the comments! Keep them coming...

Hi Josh. Re: your question:

"1. Did you control for language differences? Did all the Hong Kong students speak English as a first language? Did the American students? That's a source of potential interference, perhaps. If non-native speakers are translating into another languages with different terms or distinctions, that could pose some problems."

I think Bryce's answer is on the ball. The non-Western participants (majority) were likely not native English speakers. However, they all were taught under the education system implemented by the British (pre-handover days), where English was the second language of instruction beginning in primary school, and where English was considered extremely important for future prospects.

I should add that HKU is an English language university, with all courses taught in English. Students without fluency in English would be at a severe disadvantage in this environment.

In other words, you're right to raise this worry, but I suspect that it is not a big concern with HKU students in particular.

Also, regarding Bryce's comment:

"I was actually inclined to read the fact that we got these results from subjects in Hong Kong as pretty strong confirmation of the result as these are likely to be the most Western of any sort of East Asian participants we could have used."

I think this is also right, given former British rule.

Edouard, thanks for reminding us to re-visit our paper in light of Arico and Sytsma.

[hagop sarkissian]

oops, forgot to sign in. that was me in the comment above.

[Josh Weisberg]

Thanks, Bryce--I look forward to seeing your results if you pursue some of these questions!

Thanks Hagop.

I figured that the HK folks would be proficient English speakers, but it is something that sometimes has an influence.

Cheers,

Josh