linkedin facebook twitter rss

20 Jun Allusion of Clarity

Ambiguity to UnderstandingSpeaking to an associate at work I told him that a certain tool allows us to perform a complex technical change “with impunity”. My intent was to point out that without the tool, the change would create major ripple effects, breaking large amounts of code that would need to be fixed. He said that he would restate the advantage as “easier”. My immediate reaction was that the imprecision of “easier” is compounded by the misinterpretation that easy things to do are characterized by “impunity”, which is often untrue. But in many ways, he is right: simplifying communication leads to understanding.

Context in Understanding Dialog

Clear vision and clear speech lead to clear understanding. right? Consider the image at right. Do you see an hour glass or people speaking to one another?  The value of story-telling and the richness of metaphor and allusion in communication make me question this idea. In my estimation, many of the best authors make liberal use of metaphor, reference and allusion. George R. R. Martin, an acclaimed fantasy author, scatters references to other authors, books, lyrics and other familiar allusions throughout his writing. When I’m reading along in his manufactured worlds, and hit a reference to another character in another manufactured world, or a song or something else from popular culture, I often feel a mental jolt. Most times I like that, but sometimes it seems more forced or less natural. Whenever it happens – my brain reacts with increased activity. Exformation is stimulating.

Spoken language and written language differ in many ways, but references to shared knowledge, whether from popular culture or common experience between the speakers, are important and valuable. Stories, jokes and parables are classic examples of the meaningfulness of incorporating shared knowledge in communication. These references can often lead to multiple levels of understanding and richer mental, emotional and social experiences. These shared experiences create a common context.

Understanding Context Cross-Reference
Click on these Links to other posts and glossary/bibliography references

  

Section 5 #4

Fuzzy Logic Icon

 

Table of Context

Complex Communication Patterns

There are many communication behaviors that change the way we react to one another socially (and cognitively). These include story-telling, deception, allusions, metaphor, sarcasm and exaggeration. Despite all these complex forms of communication, we humans tend to be able to sort them out effectively, and we tend to be able to prioritize which phrases and images in popular culture are worth remembering and sharing. Because of this ability to prioritize, and our ability to think in context, friends and lovers are often able to complete one anothers’ sentences.

The presence of all these elements in our cognosphere, and our ability to sort them out, are fundamental concerns when we seek to model human intelligence. The system that understands context well enough to complete your sentences for you, may someday be able to do two really important things:

1) Recognize idioms, allusions, metaphor, sarcasm, exaggeration, deceipt and other indirect conventions of human interaction

2) Differentiate between the parts of literature, song and communication that are memorable and that are not

UntruthEven if systems can get smart enough to do all this, it is difficult to imagine systems able to interpret and derive subtext from complex body language and facial expressions. I recall a science fiction story, I’m thinking by Ursula K. Leguin, about a young woman who could ostensibly read minds because she watched people’s faces so closely that she detected even the smallest twitch. Part of this young woman’s trick was that she was able to synthesize this perceptiveness with her astute observations of human behavior, and synthesize context beyond what most less observant people could do. The timing of the twitch in the dialog and in the stream of the conversing people’s lives gave her enough clues to infer what was really going on in the speakers’ heads despite the superficial intent of their words.

Like the image of human dialog that looks like a poorly drawn hourglass or an abstract representation of vague ideas, human language is almost never an exact representation of real-world truth. We can, therefor, safely assert that all speakers and all listeners are obligated to approximate meaning. Fortunately, the mechanisms of the brain are remarkably good at approximation. We are remarkable able to add past experiences, and even fabricated fictional dialog, to derive meaning from the things we hear spoken and written.

As we explore the brain and the mechanics of understanding dialog in context, I will propose ways in which computational systems, built upon streams of ones and zeros that may seem black and white, may be able to interpret shades and metaphors that make the experience of human dialog so rich and rewarding and confusticating. How soon will these capabilities emerge in sentient computing? I don’t know. This level of advancement requires highly robust automated learning heuristics, and is likely to arrive after the end of code.

Click below to look in each Understanding Context section
 


 

 

Leave a Reply

Your email address will not be published. Required fields are marked *