The topic for today's discussion was the role of intentions in communication. The discussion focused on how speakers communicate intentions and how listeners are able to infer those intentions. We also briefly discussed the acquisition of intentional understanding.

I. Gibbs.

A. Getting the discussion rolling . . .

According to Gibbs, in order for communication to succeed, the speaker and listener have to make certain assumptions about the other's intentions. However, intention is not automatically assumed, and Gibbs outlines the four stages that lead to assuming intention. We briefly discussed the first two. First, there must be comprehension, an understanding of what the speaker says (separate from understanding their intention). This is followed by recognition, consciously identifying the intention in what is being said.

Gibbs also defines intentions as an interaction between speaker and listener rather than as something that is purely an individual phenomenon (i.e., "in the head"). We agreed that there is an interaction between the intentions and interpretations between speaker and listener, but the class also agreed that Gibbs' definition of intentions lacks some of the key characteristics of intentions (e.g., that a person can intend to do something on their own, never voice the intention and never act on it - something unique to the individual).

B. Questions and further discussion.

The first question that came up in response to Gibbs article regarded the role of conscious processing in understanding a speaker's intentions. Do people really process intentionality information at a subconscious level? What does our model gain by leaving this processing out of consciousness? Is this a necessary conclusion? If we do accept this conclusion, what else does this imply about the perception of intentions? Do some aspects of intentional understanding require more or less consciousness?

• Perhaps this system is not available to consciousness, but is that important, as long as it is carrying out its function? Is it necessary to have direct access to this process?

• Or perhaps it is available to consciousness, but is routinely automatic. By being automatic, it is not uncontrollable, but detailed cognitive processing isn't necessary.

• Conscious processing might be occurring, but it happens so briefly that is subsequently unavailable to memory.

• Conscious awareness of this system might be more important in situations where "automatic" processing would falter, such as in cross-cultural communication.

• In general, the inclusion of this section of the paper seems to push for empirical research on why we don't have direct access to comprehension of intentions.

Do we have to build up an understanding of conversation as we go along? Do we use top-down or bottom-up processing, or a combination of both?

• It seems to be a combination of top-down and bottom-up processing. If you hit a point of misunderstanding in communication, you may either go "down" (e.g., listen more closely to the sounds of the words or grammatical structure) or "up" (e.g., think about the context and what you know about communication).

• Do we always go back and reinterpret in the case of inconsistencies in communication?

• It may take a very clear violation of pragmatics to prompt a reassessment of what was said. Perhaps the mechanism for a close assessment of intentions lies in small versus larger ambiguities in what is communicated (i.e., large ambiguities assessed, small ambiguities assumed).

Is it easier to successfully communicate on the literal level or the pragmatic level? Is literal language truly transparent?

• The conventional system provides more constraints. To interpret literal meaning, you have to have a theory.

II. Grice.

A. Getting the discussion rolling (Part II) . . .

We then moved on to a discussion of Grice's ideas on meaning and intentions. Grice believes that speakers' intentions are "in their heads". The speaker first forms an intention, then communicates that intention. Grice distinguishes between two sorts of meaning. Nonnatural meanings allow you to infer that a person intentionally brought about the event (e.g., "the bell means the bus is full" indicates that the bell ringer is intentionally communicating that the bus is full). Natural meanings refer to the information that is present in the object itself (e.g., "the dark clouds mean rain").

B. Questions and further discussion: The discussion of Grice's article led to questions of the role of the speaker in expressing intentions.

How can we resolve the apparent demand on the speaker's processing? (referring to the processing involved in order to successfully communicate intentions) Does a speaker ever just express, without tailoring her communication for her audience? (e.g., blurting-is it intended to communicate anything? Is this conversation?)

• Psychoanalytically, "there are no accidents", in these incidents you are in fact trying to express something. When we violate conversational norms, it alerts people to our intentions.

Who has the upper hand in determining intentionality, the speaker or the listener? Grice would say the speaker. Gibbs would say that it "floats around in the middle".

Why do we want to "put intentions in the head"? What do we gain by thinking of intentions as only in people's minds? They do exist in people's minds, but how can we use them if they are only in the mind?

• Intentions are in people's heads. Meaning is neither "owned" by the speaker or the perceiver-it is dependent on lexical cues.

• In some cases, the listener can contribute to meaning and have some effect on the meaning of the conversation.

What is the meaning of the utterance?

• The speaker has an intention to communicate a certain meaning and produces an utterance to convey that meaning. This utterance is jointly construed by both the speaker and listener. If the listener misinterprets, the speaker's intention will have been miscommunicated.

• A distinction should be made between the meaning of the utterance, and the meaning that the speaker intended to convey. According to Grice, we should always talk about what people mean (not utterances). However, we think of words themselves as having meaning, apart from the speaker's intentions. If we only interpreted utterances by the meaning that the speaker is trying to convey, does this completely depart from the conventionality of words?

• Idioms are an example of the speaker's intentions being more important than the conventional meaning of the words; however, the idioms themselves have taken on a conventional meaning of their own. We make a distinction between what a speaker meant by a certain phrase and the literal meaning of the words. Therefore, we must be aware of the separateness of intentions and the literal meanings of words.

III. Baldwin and Tomasello.

A. Getting the discussion rolling (Part III) . . .

Finally, we discussed the Baldwin and Tomasello article. This article emphasized the importance of the speaker's communicative intentions and the perceiver's understanding of those intentions. It was shown that infants use both joint attention and shared subjectivity to understand the speaker's intentions.

B. Questions and further discussion.

Are children's abilities of intentional understanding exclusive to word learning and language orare they generalized to social cognitive and interactive abilities?

Are children really understanding intentions?

• Intentional understanding may or may not be the most parsimonious explanation for the results of these studies. It is doubtful that children have a concept of intention, but they respond to cues to intentionality that direct their attention and openness to learning-they are sensitive to what we refer to as intentions.