PARTICIPATION AND AGENCY

Here’s an excerpt of my conversation with Stewart on the D&D forums about Bleecker’s article. The discussion is basically about the limits of participation, and Stewart does a pretty good job of bringing out some of the main features of the view. Its a bit long, but it gets better as it goes. By the end I think I build up to something like a response to Kripke’s criticism of meaning, which is a result I didn’t quite expect, but I’m very happy with it.

Stewart starts by responding to my commentary on Bleecker’s article.

Under that rubric, heres some examples of other things that ‘participate’: trees, rocks, clouds, weather, Mars, clothes, brick walls, terrists, wristwatches, what

Well, Bleecker has a way of differentiating here: spimes are self-describing. They assert their presence, they make it an issue for others. Nothing you mention, except maybe terrrrritz and wristwatches, are assertive in this way.

Yesterday I was driving and I used my eyes to query what was in front of me. The brick wall downloaded information in the carrier-form of photons onto my retina describing itself as a brick wall: but not just that it was brick wall, but what color it was, how high it was, how thick it was, how old it probably was, its exact spacetime location, whether or not it was an attractive brick wall, and what the current weather was (because if it was wet it was probably raining). As such, I knew I had to turn (because it asserted its presence to me) otherwise it would have been an issue for me had I ran into it.

There isn’t a difference between being 6 feet tall, and saying “I am 6 feet tall”?

Yes there is. But since we’re not talking about ‘mentalistic’ terms the point falls through.

Well, that depends on if you think “self-describing” and “asserting” are mentalistic phenomena. Bleecker doesn’t, yet he’s willing to attribute them to animate objects.

Are you saying I can’t assert “I am 6 feet tall” unless I have a mind?

Bleecker means assertion in a trivial sense. An entity asserts if some change in undergoes causes social-effects. But like I point out, a brick wall undergoes change when it is bombarded by photons and causes people to avoid hitting it. So although I dont disagree with Bleecker’s points, we have to keep in mind he’s using handicapped terms.

I am saying a person asserting “There is a fire” and an electronic circuit powering a speaker being completed upon the presence of smoke are two entirely different things. As is the general source of our disagreements.

Where does the difference lie? Both these seem to be cases of asserting, and not in a trivial sense. It would be trivial if Bleecker claimed that the simple fact of there being fire was itself asserting that “There is fire”, but so far no one has argued for that position.

Thats why philosophy is still in business. Sure you can argue there is no difference, but that would be wrong. Completely ignoring the difference (which Bleecker doesnt do: he specifically points out hes not identifying mental attributes with what these objects do) doesnt help.

I’m arguing that the difference is further down the pole. Arbitrarily drawing the line at humanity is silly, and mistunderstands the very nature of communication and social meaning.

A person asserting is an intentional action. A fire alarm going off is a physically necessary event.

But its an event that means something. It gets me to think “hey, there’s a fire” and to move my ass outdoors. Thats different from a falling rock or a creaking branch.

The relevant difference isn’t between human communication and mere mechanical events, it is a difference between meaningful communication (social interaction, ‘blogging’ in Bleecker’s sense) and simply being-there. And if thats the relevant distinction, then the smoke detector falls on the former side. And thats accepting Bleecker’s methodological assumption of rejecting mentalistic terms.

Ultimately you have to cash it out in terms of use for a person, otherwise its utterly pointless.

Humans aren’t the only social creatures, I don’t know why you’d think otherwise. The idea is that objects can be part of our social situation in a participatory way, just like our pets can be part of our community, and neither requires that we see those objects as mere infrastructure or tools for the advancement of human beings. My dog is a part of the tribe, and should not be treated instrumentally as a minimal ethical constraint on its behavior towards it simply in virtue of the fact that it is a participatory member of the tribe. That doesn’t mean all normative considerations applicable to humans also apply to dogs, but just that it can be treated as a member, and receive the consideration that comes along with that. And the same goes for our social objects.

Yeah thats fine, but like I said, every single event imaginable falls on the former side. There is no difference between a fire alarm going off and a rock falling. The meaning lies in the human giving it meaning. A falling rock is given meaning in the same way an alarm going off. Now the fire alarm might be interrelated with a bunch of other devices through physical necessity, but meaning doesnt enter the picture until we have the person (and I dont mean to restrict ‘meaning’ to our particular species, but agents like humans).

Even ignoring my mentalistic talk of ‘meaning’; under Bleecker meaning simply reduces to having causal effects on other devices. A rock falling has effects categorically like an alarm going off and posting about it on its blog. So all the crazy philosophical things you can say about Bleecker’s world, you can say about our own. The brick wall communicates with me and has meaning for me etc.

Thats not to deny new technology changes how we interact with our environments and the rules we follow and whathaveyou, and thats interesting, but not metaphysically interesting.

Bah, meaning isn’t metaphysical either. It is social, and his point is simply that machines participate in those social practices.

On Bleecker’s model, we have a bunch of social practices already up and running by groups of humans. But once those practices are up and running, than anything capable of filling certain roles within that practice can participate. And in participating, they shift the way those practices are conducted, and the meaning attributed to them. So its tethered to human social activity, but it operates in tandem with it, as expanding its scope; not just as instruments to further uniquely human activities.

So its just looking at it in the wrong way to say “it changes our practices”, because they aren’t our practices in the first place. They are just practices. And all sorts of agents can participate in them.

So sure, lots of the things you mention could concievably participate in this deflationary sense, but the fact of the matter is that they don’t. They wind can effect change, but its not the kind of reliable agent we can depend on as a participant in the practice. I can take out the trash, I can train my dog or a robot to take out the trash, but the wind is just not competent at the task.

First of all, meaning is founded in the notion of ‘agent’. You cannot have meaning if you have no agents.

The social practices we already have up and running have meaning and value only for the agents in them. Furthermore, the entire purpose of these social practices is for the value to the enduser. THere is no point to running a business if at no point does an enduser recieve something of value. Google cashes out in terms of providing search results for its enduser. There is no ‘meaning’ or alternative purpose or teleology to google other than providing results for its user (ignoring its other services here).

You seem to believe there is some teleology to social practices not founded on an enduser (agent). Human social pratcies and technology exist only insomuch as we value them. THere is no point in replacing an entire social practice with gadgets. Theres no point in like ballroom dancing if the only things moving are robots.

You imply this interchangeability between man and machine which simply doesnt exist.

I am arguing for the interchangeability, sure. Which means you can’t go from the assumption that the interchangeability doesn’t exist to a refutation of the argument.

My argument goes like this:

1) Communicative meanings are signals exchanged by things that use meanings.
2) Whatever uses meanings are participants within the convention that gives the signal meaning.
3) Certain technological objects use communicative meaning.
4) Therefore, Certain technological objects are participants in conventional (social) practices.

I am talking specifically about communicative meaning. I think there are other kinds of meaning (like representational meaning), but I’m not worried about that. Premise 2 says that meanings aren’t intrinsic to the agent, but arise from within a social convention. The existence of the convention can be explained however you like (psychologically, socially, biologically, whatever). When the convention is up and running, the things that use the signals are said to be participants within the practice. Machines can obviously use signals, so they are participants. The wind, Mars, and the brick wall don’t use any signals- they don’t do anything- so they would be excluded from participation, relative to a particular task.

So meaning isn’t founded on agents, at least not in the robust sense of the term. But it is founded on users. To use Millikan’s terminology, it is grounded in producers and consumers of signals. And thats all I’m talking about. Just because my signal ‘Fire!’ creates sound waves that impinge on the structure of the brick wall doesn’t mean the brick wall is using that signal in any meaningful sense, because it doesn’t do anything meaningful with it. However, a robot that pulls the fire alarm and calls 911 when it hears someone yell ‘fire’ does do something meaningful, and something relevant to the practice, and it should be considered a participant.

But this isn’t teleological. Goals and ends do not need to be well-defined. We can communicate without any ends or reasons. Part of my point is that machines can collect, organize, and interpret data, and it can do this for no reason other than sheer knowledge accumulation, on the off chance that it turns out useful in the future.

I just dont buy the grounding of meaning in ‘convention’. Symbols in and of themselves have no meaning, merely causal properties. This is why computers can crank out deductive statements but cannot understand them. If there were two computers out on Pluto playing chess, I would not say there is any meaning in this system. Merely a cacaphony of causal symbol exchange and processing, no different in kind than if the computers were exchanging random commands. No different in kind than any other physical interaction in the universe.

Most people would distinguish between a collection of gadgets each running Windows and communicating via bluetooth as per the wishes of an agent; versus a collection of gadgets randomly exchanging commands. The first we would call ‘technology’, the second we wouldnt call anything. You have no way of distinguishing between these two situations.

I don’t understand how the first two sentences are compatible. Signals have meaning when they are understood within the context of a convention. A loud alarm has no intrinsic meaning, but a loud alarm made by a smoke detector, within a social practice that associates the alarm as indicating fire, comes to have meaning. The convention is what helps interpret and understand the signal; it has no interpretation apart from the convention.

I don’t think thats very controversial. The controversial point is saying that the machines use the signal, as opposed to merely causally realizing them. My point is that use itself is understood within the conventions. The user is just the communicative source of a signal. And in the case of the smoke detector, the communicative signal comes from the detector itself. Thus, it uses the signal, and thus it participates.

And by appeal to convention, I have a very easy way of distinguishing between your two cases. The latter has no associated convention with which to understand and interpret its signals, while the former does. But the former isn’t meaningful in virtue of the intentions of any agent, but because their communications are used for some practice, or to accomplish some task, or is otherwise meaningful by the lights of the convention.

Ok theres two senses of symbol here: (1) A word for example is a symbol in that it evokes a particular meaning for an agent. (2) A word is a symbol in that it has particular causal interrelations with other symbols in a formal system.

You equivocate between the two. Only under sense (1) can symbols be meaningful. Furthermore sense (1) is grounded in the notion of an agent. Sense (2) is not.

Without an agent, all we have is sense (2) symbols. Without an agent there is no difference between the group of communicating gadgets exchanging commands under (a) windows conventions, (b)under a random convention, (c)under no convention.

Only when an agent (understanding the Windows convention) comes along and interprets (a) can the symbols be said to be meaningful. They are meaningful not because they are entities in some formal system (the symbols in (b) achieve this as well), but because an agent interprets the symbols in terms of his own agency. So the agent says “I know by the Windows convention ‘x’ closes the window and thats what I want to do, so that ‘x’ has meaning for me.”

Meaning is NOT the causal properties of a given entity as defined by the convention. Meaning only exists FOR an agent.

I’m not equivocating, I’m denying that 2 is relevant here. Not all causal interactions are symbolic, after all. Two pool balls colliding is not a symbolic exchange. It is a causal exchange. Under your proposed dualism, there is no reason to qualify the difference as between two ‘formal systems’, because to apply the notion of formal system is already to bring it under and interpretation and a convention and immediately move it into the domain of type 1 interactions.

My dichotomy is simply that there are brute causal interactions, and there are meaninful interactions, where the latter is understood as certain types of interactions occuring within a community of users who regard those interactions as meaningful. The former is devoid not just of agency, but of interpretation as well- it is just the way the world turns. The latter brings in meaning in that the interactions are interpreted by a user as relevant in a practice.

If we go your route, we are left with an unfortunate and seemingly inexplicable dualism, because we have to understand the sense of ‘evoke’ in non-causal terms. But thats not because meaning is mysterious, its because the distinction you make doesn’t cut the notion at its joints.

I absolutely agree, meaning (in this communicative sense) only exists for an agent, and not in terms of the causal roles it plays. I doubt anyone could give a straightforward causal story about conventions in the sense I mean, because interpretation is a practical skill, not a formal constraint. And I am willing to conceed that the alarm as a signal has no meaning for the smoke detector. But thats not my claim- my claim is that the detector is the communicator of the signal, and what has meaning for the detector is the presence of smoke, which interprets by sounding the alarm. Sure, there’s a causal story to tell about that interpretation, but its not the causal story thats relevant to the practice. What’s relevant is that the detector is the user of the signal.