And now for something completely different. This post isn’t about arguing for some specific thesis, it is more about presenting a way of looking at the world, other people, and yourself. Even though the post is a bit lighter in subject matter, it is still something I find informative so I hope readers will find it interesting as well. It’s just an exercise in looking at social phenomenon from a Darwinian standpoint. A lot of the conclusions are common sense, but I think readers will like coming to understand the mechanisms at work behind common sense.


Section 1: Problem of Intentionality

Franz Brentano is credited with reintroducing the Scholastic notion of intentionality to philosophy. Simply put, intentionality is “aboutness.” This mountain, this river, this rock isn’t about anything, but I can think and talk about a whole range of things. To possess intentionality is to possess meaning–words mean things, thoughts mean things—but mountains, rivers, and rocks don’t mean anything. Brentano’s claim was that intentionality was the “mark of the mental,” that intentionality could not be reduced to natural non-mental processes. This, of course, set off many philosophers to do just that. If you watch this video (www.youtube.com/watch?v=GaasITFDQdg), the host (I’m not sure who it is) says:

It is sometimes said that philosophy of mind has two major problems: one of them is the problem of consciousness, the other is the problem of intentionality. I don’t know about democracy, but I think that if you were to take a vote among analytic philosophers today about the latter question, intentionality, the consensus view would be that Ruth Millikan basically solved this problem around 1984 [with the publication of Language, Thought, and Other Biological Categories].”


Section 2: Biosemantics

Millikan’s solution to the problem of intentionality is known as biosemantics or teleosemantics. Biosemantics in general refers to the study of natural meaning, as when we say that a rabbit thumping its paws on the ground means that there is a predator around, and for listening rabbits to seek cover. But specifically, biosemantics is synonymous with the theory of natural meaning put forward by Millikan. (An article on biosemantics” is available here. I am going to try to present her theory in an easy-to-understand manner, so my apologies to any Millikan purists.)


Millikan’s theory is that a sign or representation has meaning (what philosophers call intentional content or just content, for short) when several requirements are met. First, there must be a device that has been designed for its ability to produce signs. By “design” I mean that the item’s ancestors were selected for reproduction because they produced some effect. This effect is what we can say is their function to bring about, or what they were designed to do. Second, the sign that is produced must be designed by the producer to vary in accord with changes in the environment. Honeybees, for example, when they have located a source of nectar return to the hive and do a dance for the waiting bees. The mechanisms in the bee that produce this dance are designed to produce dances that correspond to the location of nectar. The wiggles and turns of the bee dance correspond to the location of the nectar relative to the sun and the hive by the semantic rules of this “language”. If the nectar had been in a different location the dance would have varied in accordance with the mapping rules of the dance. The bee dance is about the location of nectar—it possesses intentionality–and the semantics of the bee dance are capable of representing changes in its location; changes in the location of nectar can be represented by changes in the form of the dance. Finally, there must be a co-adapted audience for the sign, which Millikan calls the “consumer,” which is designed to interpret the structure of the sign for use, say, as a guide for action in retrieving the nectar. In the bee dance the perceiving bees are the consumers.


In spoken human languages the producer is the speaker, the sign is what is spoken, the consumer is the listener. The producer and consumer are adapted to one another by learning such as to understand that the spoken sound we associate with, say, the word “dog” refers to dogs. Human languages are extremely flexible in representing changes in the environment–incredibly more so that the mere bee dances that can only vary with changes in the location of nectar and the sun, and can respond to no other environmental factors.


Section 3: Pushmi-Pullyus

According to Millikan, there are three kinds of intentional signs. First, indicative signs are designed to communicate a fact about the word such as that there is a cat on the mat, that hydrogen atoms have one electron, or that Jupiter is the largest planet in the solar system. Secondly, imperative signs are directions to produce some state of affairs, such as close the door, or take out the trash. Whereas creating true beliefs is the purpose of indicative communication, creating a behavior that alters the world in some way is the purpose of imperative. Finally, there are signs that are both imperative and indicative at the same time. Millikan colorfully calls these “pushmi-pullyus” after the two-headed fictional animal from Dr. Doolittle. When a bee does a bee dance it is both an indicative indicating the location of nectar, and an imperative to perceiving bees to go get it and bring it back to the hive.


Pushmi-pullyus are the most primitive kinds of representations; more advanced representational systems (perhaps only human communication systems have evolved to split the two faces) can communicate the indicative and imperative content independently. But often human communication systems are still pushmi-pullyus. Consider, “the house is on fire.” On its face it is an indicative sign stating a fact about the world, but when actually spoken the imperative content will be encoded in the pattern of stress, emphasis, and volume of the utterance. “The house is on fire!” The urgency with which it is spoken conveys that the utterance also has imperative content: call the fire department, get a bucket of water, or get out of the house, perhaps. The imperative face works by creating a desire or emotion in the listener—fear or alarm in this case–and then having the emotion perform its function of producing an intention and then behavior.


Section 3a: A Lengthy Aside Concerning Men and Women’s Communication Styles

I was in a doctor’s office one day and there was an infomercial on the TV for a hair curler. The conversation between the female hosts went something like this:

Your hair looks awesome!

No, your hair looks awesome!

Mary’s hair looks more awesome than my hair!

No way! Your hair looks way more awesome!

This kind of inane blather makes a guy want to blow his brains out, but that is because men are more attuned to the indicative than to the emotional/imperative face of language. A perfectly acceptable male conversation can go like this:

Here comes the curveball.


Here comes the slider.


Here comes the fastball.


Millikan seems to claim that human language is either imperative or indicative, but it seems to me that human spoken language is always done via pushmi-pullyus. In “The house is on fire!” it seems pretty clear that there is both indicative and imperative content. In spoken language the urgency of the imperative content can be dialed up and down in degree with the amount of stress we put on what we are saying. In a totally flat affect, the imperative content is dialed all the way down, indicating that no behavior is required. But “and don’t do anything about it” is still an imperative.


Men take female-female communication of the sort in the infomercial I mentioned, to be inane, but that is only because we are more attuned to the indicative side. By being so attuned we are missing out that 99% of the conversation is happening on the emotional/imperative side. In speaking with such heightened emotional affect, the purpose of the conversation is to create emotions in the hearer. In saying “your hair looks awesome!” with great emotional emphasis, the purpose is to create a good feeling in the listener. (Remember, truth is the purpose of indicative language, emotion and behavior is the purpose of imperative.) The women are all emotionally pinging off of each other in an extremely complex network until they are all on the same emotional page. The words are largely just strings on which to hang the emotional content. I would guess that if of one of the hosts still thought that their hair didn’t look awesome the rest of the group would redouble their efforts to convince them that it did (truth not being the purpose of the conversation) until the whole group was on the same page. I speculate that the purpose of getting everyone on the same emotional page is to then engage in some communal behavior. It is interesting that men dial up the emotional content of their speech when communicating with women, and women dial it down when speaking to men, to meet somewhere in the middle. (Informal social situations and arguments are about 99% emotion, 1% factual :).)  When you become aware of the difference between men and women as regards emotional content of language you can’t un-see it.  In the past cultured women were educated to control the emotional timbre of their voice as to produce a mutually beneficial effect on the hearer.  Notice the emotional and tonal control Ruth Roman practices in this clip from Strangers on a Train:


Section 4: Natural Information

There is one last piece of theory I need to get out of that way, and that is the difference between natural information and representations. The best way to explain this is by example. The possession of muscles might contain the information that the bearer is strong, but muscles aren’t representations of strength; the form of muscles is designed for moving things, not to be perceived by consumers so as to indicate the location of strength. Again, smoke might contain the information that there is fire, but smoke isn’t a representation of fire because smoke isn’t produced so as to be used by a consumer to indicate the presence of fire. Smoke signals, on the other hand, are representations because the patterns of smoke produced by the signalers have been reproduced historically for use by consumers so as to communicate the content of the signal. Representations are a subset of natural information. There are causal regularities in natural information that we exploit in navigating the world, and this goes for the relationship between representation and represented as well.


We will put all this theory to work explaining self-representation in Part 2.