Lecture 16- Language evolution and structure

Question 1

What is the merge model of language evolution?

Answer 1

• Step One: A proto-language of words emerges. The problem however is that these words are Imprecise with relationships unclear and numerous to remember
• Step Two: Creating chunks + automatic routines. This means variation in words decreases and production becomes easier as it is not hard to remember.
• Step Three: Under-specification that is often specified by determining words which are most closely related, showing the function of different phrases, determining where one unit starts and stops (culturally learnt as some language have this and others don’t)

Question 2

What is meant by chunking?

Answer 2

• The brain is optimized to chunk language into phrases and clauses.
• Putting words in chunks means there is less to remember as there are associations between them

Question 3

What are the 3 steps for chunking?

Answer 3

• Select
• Attach (Merge)
• Close

Question 4

Where in the brain does chunking occur?

Answer 4

Inferior frontal gyrus

Question 5

What happens at the superior temporal gyrus?

Answer 5

Phonology

Question 6

What is different in regards to the physical structure of the brain for humans in regards to the connection between the IFG and STG

Answer 6

Higher connections are more robust

Question 7

What are some problems with the IFG being proposed as the center for chunking/syntax?

Answer 7

• It might be a result of learning, not something we bring to learning (is the brain the start?)
• It may not be language-specific (music, math + anything sequential also use this area)
• What if it’s all culturally learned?

Question 8

What is grammaticalisation?

Answer 8

• Describes the process of grammar as socially learnt
• All words come from somewhere and have a unique history (cultural) which you can trace with time
• Almost always from content words to function words

Question 9

What is the basic structure of an experiment that tests if merge is a cultural artifact?

Answer 9

• Create an artificial language
• Speaker 1 teaches it to speaker 2
• Speaker 2 teaches it to speaker 3
• Study the resulting language and how the interaction has resulted in changes to it

Question 10

What are the three ways that you might teach language?

Answer 10

• Chain= person to person
• Dyad= just two people talking
• Chain of Dyads= is a combination of the two you get the passing down from pair to pair

These are all types of interactions

Question 11

What happens when iterative learning of language in a chain occurs?

Answer 11

• Get simplified and easier to explain (grouped into amount that is easy to remember + stable)
• Lose the ability to express complex ideas

Question 12

What happens when you leave a dyad with a made up language?

Answer 12

They memorize all the words accurately so do not lose the ability to express a range of things but its fixed

Question 13

What happens when there is a chain of dyads learning a language?

Answer 13

• Perfect balance

- Cultural transmission + interaction= composability (learnable and expressive)

Question 14

How do the two theories on merger differ?

Answer 14

-Camp One: Merge, however, is not socio-cultural.
It comes from neural reorganisation, which is then
used in cultural language creation
-Camp Two: “Merge” or compositionality is also
socio-cultural. Compositionality is a result of how
language is used: in conversation and learned