Precarved Spaces of Conversation

"The social weight of all spoken words" introduces us to the idea of social weighting, or the manner in which the social representations of speech changes the ways in which we use our cognitive capacities to process that speech. This theory gives a compelling argument against the frequency theorem. I wonder how social weighting plays a role in academic vs immersive language learning for second languages. I took around one year of German language class at Stanford before living in Germany for six months. Though I knew many of the words and could speak with decent grammar, my pronunciation of the different phonemes was very different from the native speakers. Even by the end of the six months, by which time I should have adjusted my accent to sound more like a native speaker, I still sounded very much like an American. Perhaps this is a side effect of social weighting, whereupon I weighted the "learning" done in a classroom more than the "learning" I did in day-to-day life.

This article transitions nicely into the idea that the characteristics of a speaker influences how we interpret and access their words. Both visually and audibly, we make assessments of people we meet based on categorical labels we have developed in the past. Perhaps this is one abstraction that humans need in order to process the huge amounts of stimuli we are exposed to in any given moment.

The most fascinating bit of research to me in "Voice-specific effects in semantic association" was the idea that the characteristics of the speaker's voice would influence the most common word associations. Language and sound are so rich with meaning that it makes the task of getting a computer to interpret language in a similar way seem impossible. I wonder if we could isolate the sound characteristics of various kinds of speakers and have a machine "learn" how to modulate their word associations in a similar way.

"Effects of phonetically-cued talker variation on semantic encoding" was interesting to me in that the researchers were able to isolate some misunderstanding of speech as an effect of that message being inconsistent with the speaker's identity. This shows how deeply our preconceived notions influence our interactions with other people. It is as if, during each conversation, we have already pre-carved out a space in which we expect the conversation to take place. And when the words fall outside of this space, our brains have a harder time figuring out what is meant.

These three articles left me curious about the effect of social relationships on speech recognition, not necessarily in the actual understanding of the words themselves but in more metalinguistic ways. For example, would the same words uttered by someone you like have a different interpretation from the same words being uttered by someone you strongly dislike? My intuition tells me that the perceived relationship between the speaker and the listener plays a role in cognitive processing of speech.