WAYS OF KNOWING: SENSE PERCEPTION
CLASS ACTIVITY I: IDENTIFYING GEOMETRIC SHAPES AND PLASTIC ANIMALS BY TOUCH ALONE
The protocol and results tables are self explanatory for students. The results should provide them with a first hand experience of the active, exploratory nature of sense perception.
Students read the following instructions before undertaking the experiment.
While partner A’s back is turned, partner B collects the 5 shapes and selects 5 different plastic animals. Partner A, with back still turned away, puts out a hand. Partner B performs the following three tests with each shape or animal in turn.
- Without moving the object at all, press it gently into Partner A’s fingertips and hold it still for a five second count.
- Keeping the same orientation move the animal across the palm of partner A’s hand towards the fingertips slowly and gently five times.
- Allow partner A to hold the plastic animal and move it around at will in the hand for a full five second count.
Partner A must try to guess the object at each stage . Don’t be too strict: a guess of “monkey” for a Lowland Gorilla is perfectly satisfactory. Record your results in the table below. Also record “Don’t know” and wrong guesses.
CLASS ACTIVITY II: VERBATIM TRANSCRIPTIONS OF ORDINARY CONVERSATION
After working through the various sense perception class activities, especially the Yarbus eye-tracking experience, students will have discovered that, just as there are no naked perceptions, there can be no naked interpretations.
This unit explores our remarkable proclivity for speech recognition. It seems that built-in search mechanisms for points of interest and significance are as fundamental to language comprehension as they are to vision and touch.
If we know a language we scan the sounds we hear for linguistic points of interest and basic grammatical order. Speech comprehension in a familiar language seems instantaneous and is very forgiving. It has a gestalt quality.We fill in the gaps; ignore any hesitations, repetitions, silences or false starts. We forgive fragmented or otherwise eccentric grammar. We are able to decode variations in pronunciation and intonation. We infer the meaning of unfamiliar words by their context. We can presume the intended meaning of an incorrectly used word or an ambiguous word. We can navigate successfully the polysemy of language.
This activity requires access to laptops with internet access. Students will attempt an accurate, written transcription of just one minute of a decade old conversation between actress/singer Bette Midler and CEO of Disney, Michael Eisner.
Working in pairs, students focus on the portion of the interview between 01:00 and 02:00. They will soon discover that even a single minute is a long time when making accurate transcriptions. They must record exactly what is being uttered including interruptions, mistakes, repetitions, false starts, hesitations and silences. They will need to use the pause and rewind buttons repeatedly. Students use BM: and ME: to differentiate between the two speakers.
PERFORMANCE AND DISCUSSION QUESTIONS
Call on several pairs to perform the conversation based only on their transcriptions.
“What did you notice?” may be the only question necessary to begin a rich conversation about how we are able to navigate the messiness of everyday speech. The following questions may help steer the discussion and help students to stay on topic.
- Explain the difference between the sounds being uttered and what is being understood by the listener?
- How, if at all, is speech recognition like listening to music?
- What is the cocktail party effect?
- Why is it so difficult to program computers for speech recognition?