Wednesday, August 24, 2011

On Containers and Content, with a Cautionary Note to Philosophers of Mind

I was recently reminded of a short paper I abandoned work on in 2001, but for which I still have a certain amount of affection. Here it is in its entirety (slightly amended and minus a few references).

On Containers and Content, with a Cautionary Note to Philosophers of Mind

The prototypical container relation is a relation between a single item or group of countable items, and some distinct item with the approximate shape of a cylinder, box, or bag, open on no more than one side, such that the volume of the container substantially exceeds and has as a subset the volume of the items contained. (However, see below for a couple of variations.) For concreteness, we may consider the prototypical container to be a bucket and its contents to be balls.

Consider some potentially interesting features of this system:

(1.) A bucket contains a ball just in case the ball is physically inside the bucket. It does not matter how things stand outside of the bucket.

(2.) In the normal (upright, gravitational) case, it takes a certain amount of effort to get a ball into a bucket and a certain amount of effort to get it back out again.

(3.) Balls take up space. A finite bucket can only contain a limited number of non-infinitesmal balls.

(4.) Balls are clearly individuated, countable entities.

(5.) It is rarely a vague matter whether a bucket contains a ball or not.

(6.) There is typically no reason why any two balls can’t go in the same bucket or why a ball can’t be removed from one bucket and put into another without changing any of the other contents.

(7.) A bucket can contain many balls, or only one ball, or no balls.

(8.) The ball and the bucket are distinct. The ball is not, for example, a state or configuration of the bucket.

(9.) A ball in a bucket is observable only from a privileged position inside or above the bucket.

Some modifications, particularly to (4) and (5), are required if we take as prototypical the relation between such a container and a certain amount of stuff, such as water or sand, characterized non-countably by means of a mass noun. If only one kind of stuff is to fill the bucket, there will be no multiple, discrete contents, but one content in varying amounts. Alternatively, if one fills the bucket with wholly distinct kinds of stuff, (4) may be preserved; if we consider semi-distinct fluids, such as orange juice and apple juice, (4) must be discarded.

If one takes the relationship between a packaging box and the packaged item as prototypical, the volume of the item contained will approach the volume of the container, requiring the modification of (6) and (7).

Cautionary note to philosophers of mind:

Often it is said that beliefs, desires, etc. -- the "propositional attitudes" -- are “contents” of minds. Also, and quite differently, propositional attitudes are said to have contents, propositional, conceptual, or otherwise. It is infelicitous to invoke the container metaphor in this way (or, alternatively, to extend literal usage of ‘content’ to cover these cases), if there are divergences between the features described above and features of the mind-propositional attitude relation or the propositional attitude-proposition relation, and if incautious use of the metaphor might draw the reader (or the writer) mistakenly to attribute to the latter relations features of the former. Similar remarks apply to visual (and other) images, which are sometimes described as contained in the mind and sometimes described as themselves having contents.

I will leave the explicit comparisons to the reader; they should be obvious enough (e.g., (1) makes “content externalism” an oxymoron; (4) - (7) are atomistic). If we give ourselves completely over to the container metaphor, we end up with a position that looks something like a caricature (not an exact portrait) of Jerry Fodor’s views. The metaphor thus pulls in Fodor’s direction, and Fodor, perhaps sensing this, delightedly embellishes it with his talk of “belief boxes.” Conversely, those among us who wish to resist the approach to the mind that the container metaphor suggests may do well to be wary the word ‘content’ in its current philosophical uses. Would we lapse so easily into atomistic habits if instead of saying that someone has (in her mind) a particular belief with the content P, we said that she matches (to some extent) the profile for believing that P?

Admittedly, avoiding the word ‘content’ would make some things harder to say – but maybe those things should be harder to say.


Kevin Connolly said...

Susanna Siegel argues that a newspaper story is a better model for content than a bucket (for some of the same reasons that you mention):

Erol Copelj said...

Phenomenologists have often emphasized the difference between "external objects" (or better: constituted objects) and intentional states via which such objects are constituted (forget idealism in this context- constituted just means: how conscioussnes presents objects). For instance, the former, but not the latter, can take spatial predicates. Husserl, in the his lecture on time consciousness for example, talks about the inherent differences between constituted objects and constituing acts. It seems to me that, as Sartre states, our thinking about consciousness is often distorted because we impose on intentional states what we observe in their objects.

In essence, I think that you may be heading for a simillar conclusion...

Clayton Littlejohn said...

Hi Eric,

I wonder how prevalent the container metaphor really is. Is it your view that talking about content is itself some indication that you buy into this metaphor? (I suppose I never would have thought about attributing that view to anyone, but maybe that's because I had already accepted the points you were trying to make and charity kicked in.)

I don't wish to defend the bucket view, but I didn't quite see the tension between (1) and content externalism. If I'm at the casino and have a bucket of coins, certainly whether the coins are in the bucket depends upon how things are inside the bucket. Whether these things are coins as opposed to hunks of metal with no value arranged like coins, however, certainly depends upon where the coins came from. So, even if you thought (which you shouldn't!) that beliefs were particular that could be held in a container, it doesn't follow that what they are is determined wholly by what's in the container.

Eric Schwitzgebel said...

Thanks for the comments, folks!

@ Kevin: Thanks for the tip on the Siegel! It's interesting to see her acknowledgement of the weakness of the bucket prototype. I suspect she might not be quite as nervous about the word "content" as I am, though.

@ Erol: I agree that it is easy to confuse properties of objects (e.g., blueness, saltiness) with properties of the experiences we have of those objects (e.g., blue*ness, salty*ness) -- not so sure about spatiality in particular, though (that's something I've been worrying about recently). I'm not sure the container metaphor is to blame here, though.

@ Clayton: You have hit on the reason why I abandoned this paper. It's an open question to what extent "dead" metaphors influence thinking. If they're dead enough, perhaps not very much. It's very hard to evaluate the extent to which ordinarily philosophical use of the word "content" does in fact create pressures toward cartoon-Fodorian views. It would be cool if there were some way to evaluate that, but I don't see anything plausible. On your second comment: That's a nice point about the coins. I was thinking of balls, not coins, as the prototypical contents. As one varies the prototypical contents or other features of the prototype, the tacit philosophical implications do shift.

Kevin Connolly said...

Hi Eric,
I don’t think that it’s the word “content” that is problematic. There is a sense of the word “content” found in phrases like “the content of a newspaper article.” That notion of content appears to avoid the problems that you raise for the bucket notion of content (content in the newspaper article sense is non-spatial, for instance). This makes a newspaper article better than a bucket as a model for mental content. But also, it seems like philosophers of mind tend to use the word "content" in the newspaper article sense rather than the bucket sense. So, I don’t think that your article shows that philosophers of mind should avoid the word “content.” It shows that they should avoid using it in one particular sense (but I’m not sure that they ever used it in that sense anyway).

David M Gray said...

Well, the debate between internalists and externalists about content is often referred to as a debate concerning the locality of mental content. This does give the notion of content a spatial element.

Regardless of the benefits of the newspaper metaphor vs. the bucket metaphor, I'm not sure the bucket metaphor makes ‘content externalism’ an oxymoron. You mention that it is often said that propositional attitudes are taken to be contents of the mind. If the mind is the container in question, then I think I agree that ‘content externalism’ would be an oxymoron.

I think I’m a bit more familiar with content being talked about as what is expressed by a that-clause. So we can talk about the contents of a mental state (e.g. that water is wet) regardless of what attitudinal relationship one has to the content (say, a belief or desire; although, this would admittedly be a strange desire to have). So the mental content is in this case content of a mental state and not content of a mind. (Here the bucket metaphor might also break down as the mental content both partially constitutes the mental state)

I’ve always thought of externalism as signaling something that is at least partially external to the head, not the mind (as in ‘meanings ain’t in the head’). While Putnam’s phrase concerns semantic externalism I think the same holds for content externalism. If so, ‘externalism’ would be picking out the relationship between mental contents and one kind of container (e.g. the head), but what make mental content a kind of content is the relationship it has to the mental states it partially constitutes or occurs 'in'.

Eric Schwitzgebel said...

Kevin: But I'm not so sure the newspaper content use isn't itself metaphorical, in which case it's not so much an alternative as another instance of the same thing. Sorry I wasn't clear about that.

David: Right, the issue plays out differently depending on whether the mental states are the contents of the mind, or have contents, or both. And I also agree that if the mind extends outside the head, contents can be external to the head without being external to the mind. Susan Hurley is good on this, I think, where she points out the difference between "vehicle externalism" and "content externalism". Many content externalists will deny vehicle externalism, saying that mental states are in the head even if their content is contingent on external affairs (as in Clayton's coins/cup example above). On most ways of slicing it, though, buckets and balls don't seem to get things quite right.

David M Gray said...


I haven't looked at Hurley on this but thanks very much for the reference! Sounds useful.

J.Vlasits said...

Even if your cautionary note is not totally relevant to philosophers of mind, it may be useful for (formal) epistemologists, who often run into trouble with, e.g. logical omniscience, because they have an implicit picture of an agents beliefs as just a set of possible worlds that work rather like balls in a bucket!

Eric Schwitzgebel said...

J.V.: Yes, interesting thought!

Jackson said...

"Ideas" are one type of "content" that has been identified as being located "within the mind." Other types of alleged "mental contents" include "emotions," "memories," "dreams," "hallucinations," and even so-called "physical sensations." However, according to Orteguian metaphysics all of these different types of phenomena are "occurrences" that are not "I," "me," the individual human person to whom the phenomena "occur." As such, all such "occurrences" are "external" to the "I," or to "me." However, both the "I," and the phenomena that occur to "me" (which Ortega calls "my circumstance"), are "internal to" or "contents of" the "radical reality" that is "my life."
[By the way, as any golfer or movie-goer knows, a bucket can "contain" (in the sense of "keep from spilling over") more that would technically be considered to be "inside" the bucket.]

nanci said...

Hi eric....My name is nanci. I was reading some of your articles and was hoping to find one on " seeing life in the dark" or something that could explain why when I turn off the lights I can see alot of strange images floating all around the room and no matter where I look they are there....If I stop and focus of the images they become real clear and its freaky.... one time I seen like an old cafee where people are being served on roller skates and like a old diner with red seats and a juke box and the the bar tender cleaning off the table area around him.....another time I seen three people sitting in front of a tv and one person kept on getting up and changing the was one of those older tv's where you turn the channel with a knob and the antena was the bunny ears....none of this I'm alseep, I mean I could close my eyes and talk you through what I am seeing as I see it. One time I asked my friend do you see those immages right there? and he said NO, your trippin...your seeing things...I said "I KNOW" I was just hoping you could see them If you have an article on stuff like that can you tell me where I could find it so I can go and read about it thanks

Eric Schwitzgebel said...

Check out the comments on this post!