The Piraha people do not have words for exact counting, and their language does not have words for numbers. This chapter discusses the absence of words for counting in the Piraha language.

The Piraha Language

An introduction to Edward Gibson, a psycholinguistics professor at MIT. This chapter provides an overview of his work and upcoming book.

Introduction to Edward Gibson

Edward Gibson discusses his early fascination with human language and how he transitioned from mathematics and computer science to the study of language.

Fascination with Human Language

Edward Gibson shares his perspective on language as a puzzle to be solved from an engineering perspective, and his initial interest in computational linguistics.

Approaching Language as an Engineer

A discussion on the gap between form and meaning in human language, and the challenges of understanding the structure of thought and communication.

The Divide Between Form and Meaning

Edward Gibson explains the generalizations in human languages, particularly focusing on word order and prepositions in different languages.

Generalizations in Human Language

An exploration of the structures of language, including sounds, words, and grammar, and the concept of dependencies between words in a sentence.

Dependency Structures in Language

A discussion on the process of constructing a tree from a sentence, including the root of the tree and the dependencies between words.

Constructing a Tree from a Sentence

Insights into the evolution of language, morphology, and the development of color words in different cultures, highlighting the functional aspects of language evolution.

Evolution of Language and Morphology

Discussion on the evolution and change of languages, including the impact of writing systems and the evolution of language on the internet.

Language Evolution and Change

Exploration of syntax, grammar, and the differences between phrase structure grammar and dependency grammar.

Syntax and Grammar

Explanation of movement in language, focusing on the theory proposed by Noam Chomsky and the alternative theory of lexical copying.

Movement in Language

Discussion on formal language theory, including context-free and regular languages, and their relevance to human languages.

Formal Language Theory

Comparison of the methods used in the study of language, including the disagreement between dependency grammar and phrase structure grammar.

Disagreement and Methodology

Discussion on the length of dependencies in language and how it affects comprehension and production.

Dependency Length and Language Difficulty

Exploration of center embedding and nesting in language, and their impact on comprehension and production.

Center Embedding and Nesting

Discussion on behavioral consequences of long-distance connections and the use of experimental methods to study comprehension and production difficulties.

Behavioral Consequences and Experimental Methods

Discussion on the completion of central embedded materials and the challenges associated with this task.

Completing Central Embedded Materials

Examination of the universal theory of language and the measurement of dependency lengths in various languages.

Universal Theory of Language and Dependency Length

Exploration of the language network in the brain and its relationship to thinking and comprehension.

Language Network and Thinking

Discussion on the relationship between language and thinking, suggesting that language is not necessary for thinking.

Language and Thinking

Exploration of large language models and their ability to predict and understand English, with a focus on the limitations of current theories.

Large Language Models

Explanation of dependency grammar and its potential role in construction-based theories of language.

Dependency Grammar

Debate on whether large language models truly understand language, with examples of their limitations in understanding meaning.

Understanding Language

Discussion on the cognitive cost associated with long dependencies in language and its implications for comprehension and recall.

Cognitive Cost of Language

Investigation into the complexity of legalese language, particularly the use of center embedding and its impact on comprehension.

Legalese and Center Embedding

Discussion on the relationship between language and communication, the concept of noisy channels, and the work of Claude Shannon in communication theory.

Language and Communication

Exploration of language optimization for communication, the role of word order and syntax, and the impact of noisy channels on language structure.

Optimizing Language for Communication

Debate on the innateness of language, the modularization of language areas in the brain, and the connection between language and thought.

Innateness of Language

Examination of language and culture, including the study of remote cultures and their impact on language, as well as the influence of culture on language development.

Language and Culture

Discussing the ability of the Pirahã people to perform matching tasks and the limitations they face due to the lack of counting words.

Learning from the Pirahã

Exploring the challenges faced by the Pirahã people in tasks involving sets and their inability to count due to the absence of counting words.

Encoding Sets

Examining the impact of language on the abilities and limitations of the Pirahã people, particularly in relation to counting and task performance.

Language as a Limiter

Discussing the influence of language on culture and the hypothesis of language development in relation to farming and counting systems.

The Role of Language in Culture

Exploring the reasons behind language diversity and language death, emphasizing the role of language in community and economic value.

Language and Community

Contemplating the potential for communication with non-human species and the similarities and differences in communication systems across species.

Communication Across Species

# The Evolution of Language: A Conversation with Edward Gibson

In this newsletter, we continue our fascinating conversation with Edward Gibson, a psycholinguistics professor at MIT, as he discusses the evolution of language and the relationship between cultural language and how people represent, process, and learn language.

## The Piraha Language

- The Piraha people do not have words for exact counting, including not having a word for "1" in their language.
- This lack of exact counting words challenges traditional notions of language and communication.

## Edward Gibson's Journey

- Edward Gibson's fascination with human language began as a child in school when he found the process of structuring sentences in English grammar interesting and confusing.
- Initially a math and computer science enthusiast, Gibson stumbled upon language as a puzzle to solve from an engineering perspective, which eventually led him to a career in computational linguistics.

## Word Order and Meaning

- Gibson delves into the fascinating generalizations that occur across different languages, such as the relationship between word order and the structure of communication.
- He explains the concept of dependencies in language and how all sentences can be broken down into a tree structure, with each word depending on another.

## Morphemes and Evolution of Language

- The discussion touches upon the concept of morphemes, the minimal meaning units within a language, and how different languages have varying levels of morphology.
- The evolution of language, including the development of color words, is explored as a function of communication and the needs of early tribes.

## Syntax and Grammar

- Gibson shares insights into syntax and grammar, discussing the differences between phrase structure grammar and dependency grammar.
- He highlights the disagreement between Noam Chomsky and himself regarding the movement theory and the lexical copying theory, shedding light on the history of the study of language.

## Language and Thought

- The conversation delves into the relationship between language, thought, and the human brain.
- Gibson discusses the separation between language comprehension and thinking, highlighting how the language network in the brain is distinct from other cognitive functions.

## The Limits of Large Language Models

- Gibson and his colleague, Eric Martinez, conducted a study on legalese, the language used in contracts and laws, and found that it is characterized by a high frequency of center-embedded clauses, which significantly impacts comprehension and recall.
- They discovered that even lawyers, who are accustomed to legalese, prefer the less complicated, un-center-embedded versions of legal texts.

## Noisy Channels and Communication

- Gibson discusses the concept of noisy channels in communication, drawing from the work of Claude Shannon in information theory.
- He explores the impact of background noise, speaker inherent noise, and listener inherent noise on language communication and the optimization of passing messages from one side to the other.

## The Challenge of Communicating with Other Species

- Gibson shares his thoughts on the potential for communication with intelligent alien civilizations and the challenges of understanding their way of communicating.
- He draws parallels between this concept and his experiences in learning and understanding the languages of remote tribes.

## Conclusion

The conversation with Edward Gibson provides a thought-provoking insight into the evolution of language, from the structure of sentences to the development of vocabulary and the impact of cultural and societal influences.

Stay tuned for more fascinating insights into the world of language and cognitive science. And don't forget to check out our sponsors in the description to support the Lex Friedman Podcast.

Unraveling the Evolution of Language with Edward Gibson

The subject comes first, followed by the verb, and then the object

The verb comes first, followed by the subject, and then the object

The object comes first, followed by the verb, and then the subject

The object comes first, followed by the subject, and then the verb

Form is like the shadow of the bigger meaning underlying language

Form and meaning are completely independent

A rule that allows words to move from one structure to another

A rule that generates new words with slightly different argument structures

A rule that allows words to copy their meanings from other words

A rule that determines the ordering of words in a sentence

It leads to problems in understanding syntax

It leads to difficulties in language production

It leads to challenges in word recognition

The movement story makes the learning problem easier

The movement story makes the learning problem impossible

The movement story leads to difficulties in language learning

The movement story has no impact on the learning problem

The speaker gathers data from experiments and corpus analyses

The speaker focuses on thought experiments and intuitions

The speaker is more interested in philosophy and psychology

The speaker evaluates theories quantitatively

Language is a communication system, thought is meaning

Language is thinking, thought is communication

Language is verbal, thought is non-verbal

Language is structured, thought is unstructured

During the first year of life, as the brain begins to develop

It's hard to get a 2 year old to do this task, but 3 and 4 year olds are starting to do it for short periods

Language development occurs primarily during sleep

Language development is not influenced by age

High-level comprehension of language, including spoken and written language

Low-level comprehension of language, including spoken and written language

The language network is not activated during production

The same network is activated as in comprehension of language

The network for production is different from the network for comprehension

Production of language is easier than comprehension of language

Language and thinking are separate systems

Thinking is dependent on language comprehension

Language is the primary driver of thinking

The cost of producing and comprehending connections between words which are not beside each other

The cost of understanding complex sentences

The cost of processing low-frequency words

It has a performative meaning that is associated with the style of legalese

It simplifies the language for better comprehension

It reduces the cognitive cost of longer distance connections

It conforms to the rules of dependency grammar

Lawyers intentionally make legal language hard to understand for financial gain

Lawyers intentionally make legal language hard to understand as a performative aspect

Lawyers unintentionally make legal language hard to understand

Lawyers have no incentive to make legal language hard to understand

Language has no relationship with thought

Western educated, industrialized, rich, and democratic people

Isolate languages with no known connected languages

Lack of mathematical basis for translation

Difference between Dostoevsky and Tolstoy

Difference between Hemingway and Bukowski

- The Piraha people do not have words for exact counting, not even for the number 1
- Edward Gibson, a psycholinguistics professor at MIT, discusses his fascination with human language
- He became interested in language as an engineering puzzle and studied computational linguistics at Cambridge
- He finds beauty in the generalizations and patterns of word order in different languages
- He believes that word order in language is influenced by minimizing dependencies between words.

- The speaker discusses the relationship between word order and sentence structure in different languages
- They explain the concept of dependencies in language, which refers to the connections between words in a sentence
- The speaker uses the example of constructing a tree to illustrate how words are connected in a sentence
- They also discuss the flexibility of word order in different languages, such as Russian
- The concept of morphemes, which are the minimal meaning units in a language, is also explained
- The speaker suggests that the process of forming a tree in language is automatable and can be analyzed using linguistic principles

- The video discusses the evolution of language and morphology, focusing on the past tense and color words in different languages.
- It explains the concept of morphemes and how they are used to convey meaning in words.
- The discussion also touches on the evolution of language and the factors that influence the development of language, such as communication needs and cultural influences.
- The example of color words in different cultures is used to illustrate how language evolves based on functional and aesthetic needs.
- The lack of data on the evolution of language is highlighted, with the example of old English to modern English being one of the few examples of language evolution that can be studied.

- The piece discusses the evolution of language from Old English to Modern English and the lack of writing systems in many languages.
- It also touches on the evolution of language on the internet and the influence of humor and slang.
- Noam Chomsky's contributions to formal language theory and the differences between phrase structure grammar and dependency grammar are also discussed.
- The concept of movement in language, as proposed by Chomsky, is explained as well.

- The piece discusses two theories of grammar: one with movement and one with lexical copying
- The movement theory proposes that words and phrases move from one structure to another, while the lexical copying theory suggests that words have two forms with different argument structures
- The lexical copying theory is seen as more advantageous for language learnability and predicts that not all auxiliaries might move
- The piece also touches on the distinction between phrase structure grammar and dependency grammar, with a focus on the lengths of dependencies between words
- Chomsky's work on context-free and regular languages is mentioned, with the assertion that human languages are at least context-free and possibly more complex
- The piece also discusses the limitations of regular languages in capturing long distance dependencies and recursion, which are present in human languages.

- The piece discusses Noam Chomsky's theories on human language and the disagreements between him and David Poeppel.
- Chomsky proposed theories in 1957 and 1965 based on phrase structure and movement, but later realized they led to learning problems and suggested language is innate.
- Poeppel disagrees with Chomsky's methods, as he gathers data from experiments and corpus analyses, while Chomsky relies on thought experiments and intuitions.
- Poeppel prefers the dependency grammar framework for understanding language, as it represents the distance between words and their connections in a more transparent way.

- The speaker discusses the complexity of language and how nesting dependencies in sentences can make them difficult to understand in any language
- They explain that long-distance connections between dependents make sentences hard to produce and understand
- Various experiments have been conducted to measure the difficulty of understanding and completing sentences with nested dependencies
- The universal theory of language by Ted Gibson states that all languages have short dependencies, and this has been supported by research on 60 languages
- The length of dependencies in real languages is always shorter than in a control scenario, regardless of word order or language type

The Piraha people lack words for exact counting, and Edward Gibson discusses his fascination with human language. He explores word order and dependencies in language, as well as the evolution of language and Noam Chomsky's theories. The complexity of language and nesting dependencies in sentences is also discussed, along with the universal theory of language proposed by Ted Gibson.

- The discussion revolves around the evolution of language and the purpose of language in terms of communication and processing.
- There is a debate about whether language evolution is optimized for ease of production or for the listener's understanding.
- The connection between language and thought is also explored, with the stability of the language network in the brain being a focal point.

- The piece discusses the challenges of conducting fMRI scans on young children to study language development.
- It highlights the difficulty in getting 2-year-olds to participate in the task, but shows that language development starts early.
- The language network in the brain is activated for speech, listening, and reading, but not for non-language tasks like programming.
- The piece also explores the concept of an inner voice and the separation of language comprehension from thinking.
- It mentions a study on patients with left-sided strokes who lose their language abilities but retain other cognitive functions.

- Language is not necessary for thinking, as shown by patients who can do math without language abilities
- Large language models (LLMs) are the best at predicting English but lack detail and are opaque
- LLMs may be implementing something like dependency grammar, which is a form of construction-based theories of language
- LLMs are good at mimicking language form but do not understand deeper meanings and can be easily fooled
- LLMs make errors that humans do not, suggesting a difference in understanding and mechanism of error detection

- The discussion focuses on the limitations and similarities between large language models and human language processing
- The cognitive cost associated with longer distance connections in language is discussed, with a focus on dependency grammar
- The idea that words close together have a lower cognitive cost is explored, and there is interest in developing a mathematical formula to explain this phenomenon

- The speaker discusses the complexity of legal language, known as legalese, and how it affects comprehension.
- They conducted a study on contracts and found that legalese is characterized by center embedding and low-frequency words, which hinder understanding and recall.
- The study involved both laypeople and lawyers, with similar results showing difficulty in understanding and recalling legalese.
- The speaker suggests that the use of center embedding in legalese may serve a performative function, similar to a "magic spell" to convey certainty in legal language.

- The piece discusses the complexity of legal contracts and the incentive for lawyers to make them hard to understand
- It explores the concept of "noisy channels" in communication and how language may have evolved to optimize communication in the presence of noise
- The work of Claude Shannon in communication theory and his interest in language is also mentioned
- The idea that word order in language may be optimized for robust communication in noisy environments is discussed
- The piece concludes with a discussion on the different sources of noise in communication and the need for a better model of noisy channels.

- Language evolution and its purpose in communication and processing are discussed
- Debate over whether language evolution is optimized for production or listener's understanding
- Connection between language and thought, stability of language network in brain
- Challenges of fMRI scans on young children for language development study
- Language network in brain activated for speech, listening, and reading, not for non-language tasks
- Exploration of inner voice, separation of language comprehension from thinking
- Study on patients with left-sided strokes losing language abilities but retaining cognitive functions
- Large language models (LLMs) limitations in predicting English, lack of understanding and easily fooled
- Cognitive cost of longer distance connections in language, focus on dependency grammar
- Complexity of legal language, impact on comprehension and recall
- Use of center embedding in legalese for performative function
- Language evolution for robust communication in noisy environments, Claude Shannon's work mentioned
- Word order optimization for robust communication in noisy environments
- Need for a better model of noisy channels in communication.

- The speaker discusses various factors that affect language learning, such as age, language form, and speech impediments
- They question the correlation between language structure and ease of learning
- They debate the innate nature of language and its relationship to thought
- They mention natural experiments involving brain scans of individuals with language deficits
- They suggest that language development may occur separately from thought, contrary to Chomsky's belief

- The speaker discusses the separation of language and thought, and the use of fMRI to measure language networks in the brain
- They emphasize the importance of studying non-industrialized cultures to understand the connection between language and culture
- The speaker has worked with the Chimani and Piraha groups in the Amazon, both of which have isolate languages
- The Piraha language lacks specific words for numbers, and their quantifiers are approximate and context-dependent
- This challenges the idea that all humans have words for exact counting

- The Pirahã people can perform matching tasks without counting
- They struggle with tasks that require counting or exact numbers
- The development of counting may be linked to farming and the need to keep track of livestock
- Language survival is tied to economic value and national identity
- There is a tension between convenience and trade and the preservation of cultural and national identity through language

- The speaker discusses the challenges of machine translation and the cultural connections to language
- He mentions the difficulty of translating concepts that don't exist in other languages
- He talks about the potential for communication with animals and plants
- The speaker emphasizes the importance of intellectual humility in understanding different forms of communication
- He discusses the possibility of communicating with intelligent alien civilizations
- The process of learning a new language is also mentioned, emphasizing the importance of social interaction
- The speaker advises young people to pursue their interests and be open to new experiences.

- The speaker had trouble visiting Pio do Ha in Brazil due to corrupt government officials
- They then found the Chamani and were able to visit with limited contact
- The speaker enjoys trying new things and pursuing unique experiences
- The conversation ends with a quote from Wittgenstein about the limits of language and the world

The speaker discusses factors affecting language learning, debates the innate nature of language, and emphasizes the importance of studying non-industrialized cultures. They also discuss the challenges of machine translation, the potential for communication with animals and plants, and the importance of intellectual humility. The speaker advises young people to pursue their interests and be open to new experiences. They also share their experiences visiting the Chimani and Piraha groups in the Amazon and end with a quote from Wittgenstein about the limits of language and the world.

Edward Gibson is a psycholinguistics professor at MIT and heads the MIT Language Lab. Please support this podcast by checking out our sponsors:
- Yahoo Finance: https://yahoofinance.com
- Listening: https://listening.com/lex and use code LEX to get one month free
- Policygenius: https://policygenius.com/lex
- Shopify: https://shopify.com/lex to get $1 per month trial
- Eight Sleep: https://eightsleep.com/lex to get special savings

TRANSCRIPT:
https://lexfridman.com/edward-gibson-transcript

EPISODE LINKS:
Edward's X: https://x.com/LanguageMIT
TedLab: https://tedlab.mit.edu/
Edward's Google Scholar: https://scholar.google.com/citations?user=4FsWE64AAAAJ
TedLab's YouTube: https://youtube.com/@Tedlab-MIT

PODCAST INFO:
Podcast website: https://lexfridman.com/podcast
Apple Podcasts: https://apple.co/2lwqZIr
Spotify: https://spoti.fi/2nEwCF8
RSS: https://lexfridman.com/feed/podcast/
Full episodes playlist: https://www.youtube.com/playlist?list=PLrAXtmErZgOdP_8GztsuKi9nrraNbKKp4
Clips playlist: https://www.youtube.com/playlist?list=PLrAXtmErZgOeciFP3CBCIEElOJeitOr41

OUTLINE:
0:00 - Introduction
1:13 - Human language
5:19 - Generalizations in language
11:06 - Dependency grammar
21:05 - Morphology
29:40 - Evolution of languages
33:00 - Noam Chomsky
1:17:06 - Thinking and language
1:30:36 - LLMs
1:43:35 - Center embedding
2:10:02 - Learning a new language
2:13:54 - Nature vs nurture
2:20:30 - Culture and language
2:34:58 - Universal language
2:39:21 - Language translation
2:42:36 - Animal communication

SOCIAL:
- Twitter: https://twitter.com/lexfridman
- LinkedIn: https://www.linkedin.com/in/lexfridman
- Facebook: https://www.facebook.com/lexfridman
- Instagram: https://www.instagram.com/lexfridman
- Medium: https://medium.com/@lexfridman
- Reddit: https://reddit.com/r/lexfridman
- Support on Patreon: https://www.patreon.com/lexfridman

Naively, I certainly thought that all humans would have words for exact counting, and the Piraha don't.

There's not a word for 1 in their language.

And so there's certainly not a word for 2, 3, or 4.

So that kind of blows people's minds off.

How are you gonna ask, I want 2 of those?

And so that's just not a thing you can possibly ask in the Pewterhau.

The following is a conversation with Edward Gibson, or Ted, as everybody calls him.

He is a psycholinguistics professor at MIT.

He heads the MIT Language Lab that investigates why human languages look the way they do, the relationship between cultural language and how people represent, process, and learn language.

Also, he should have a book titled Syntax, A Cognitive Approach, published by MIT Press coming out this fall.

To support it, please check out our sponsors in the description.

And now, dear friends, here's Edward Gibson.

When did you first become fascinated with human language?

As a kid in school, when we had to structure sentences in English grammar, I found that process interesting.

I found it confusing as to what it was I was told to do.

I didn't understand what the theory was behind it, but I found it very interesting.

So when you look at grammar, you're almost thinking about like a puzzle, like almost like a mathematical puzzle.

I didn't know I was gonna work on this at all at that point.

I was really just, I was kind of a math geek person, computer scientist, I really liked computer science.

And then I found language as a neat puzzle to work on from an engineering perspective, actually.

That's what I, as a, I sort of accidentally, I decided after I finished my undergraduate degree, which was computer science and math in Canada in Queens University, I decided to go to grad school.

It's like, that's what I always thought I would do.

And I went to Cambridge where they had a master's in, a master's program in computational linguistics.

And I hadn't taken a single language class before.

All I'd taken was CS, computer science, math classes, pretty much, mostly, as an undergrad.

And I just thought this was an interesting thing to do for a year, because it was a single year program.

And then I ended up spending my whole life doing it.

So fundamentally your journey through life was 1 of a mathematician and a computer scientist.

And then you kind of discovered the puzzle, the problem of language, and approached it from that angle, to try to understand it from that angle, almost like a mathematician or maybe even an engineer.

As an engineer, I'd say, I mean, to be frank, I had taken an AI class, I guess it was 83 or 84, 85, somewhere 84 in there a long time ago, and there was a natural language section in there, and it didn't impress me.

I thought there must be more interesting things we can do.

It didn't seem very, it seemed just a bunch of hacks to me.

It didn't seem like a real theory of things in any way.

And so I just thought this seemed like an interesting area where there wasn't enough good work.

Did you ever come across the philosophy angle of logic?

So if you think about the 80s with AI, the expert systems where you try to kind of maybe sidestep the poetry of language and some of the syntax and the grammar and all that kind of stuff and go to the underlying meaning that language is trying to communicate and try to somehow compress that in a computer representable way.

Do you ever come across that in your studies?

I mean, I probably did, but I wasn't as interested in it.

I was trying to do the easier problems first, the ones I could, thought maybe were handleable, which seems like the syntax is easier, like which is just the forms as opposed to the meaning.

Like when you're starting to talk about the meaning, that's a very hard problem, and it still is a really, really hard problem.

But the forms is easier, and so I thought at least figuring out the forms of human language, which sounds really hard, but is actually maybe more tractable.

So it's interesting, you think there is a big divide, there's a gap, there's a distance between form and meaning.

Because that's a question you have discussed a lot with LLMs, because they're damn good at form.

Yeah, I think that's what they're good at, is form.

Exactly, and that's why they're good, because they can do form, meaning's hard.

How close form and meaning are, but we'll discuss it.

But to me, studying form, maybe it's a romantic notion, gives you, form is like the shadow of the bigger meaning thing underlying language.

We communicate with each other using language.

So in understanding the structure of that communication, I think you start to understand the structure of thought and the structure of meaning behind those thoughts and communication to me.

What do you find most beautiful about human language?

See all Lex Fridman transcripts on Youtube

Edward Gibson: Human Language, Psycholinguistics, Syntax, Grammar & LLMs | Lex Fridman Podcast #426