Language as a Window into Evolutionary Psychology

Language as a Window into Evolutionary Psychology

A Jason Wilkes Production

My Teaching Style“Education can, and should, be dangerous.” –Howard Zinn

1) Please feel free to shout out questions. No need to raise hand.

2) Feel free to question or forcefully challenge what I say.

3) Any questions? Want to study for final? Want dirt on math profs?Better → [email protected]

CAB Office #666

mailto:[email protected]

Big picture – What this lecture is

● Not meant to be a summary of different theories of how language evolved.

● Dr. Snyder's slides do this well.

● If I was doing that, no reason for me to talk. Even where lectures overlap, very different.

● My lecture will use language as a case study to examine some of the fundamental issues in evolutionary psychology.

Why start with “pre-debauchery”?● Famous William James Quote

● May seem off-topic, but essential. “Fundamentalism”

● Preheat : Cooking :: Pre-debauch : My lecture● Starting on-topic only reasonable if we all share

the same premises (philosophy problem)

● A few examples to illustrate my premises, and the premises implied by modern science, so that we can understand each other later on.

Pre-debaucheryAs psychologists, our objects of study are

humans. It is therefore vital that we understand, objectively, what humans are

Pre-debauchery #1: Size

“We're an immensely large, planet-sized object.” –John Tooby


http://www.youtube.com/watch?v=LyzIau5dBao● Ending of “Men in Black” (0:21 – end)

● Thinking objectively: Should feel like encountering those aliens every time we sit next to some human on the bus.

● Ask a biochemist how large a single cell is.

● We are 100,000,000,000,000 cells.

http://www.youtube.com/watch?v=LyzIau5dBao

What we are, what each of us is, what you are, what I am, is approximately 100 trillion little cellular robots. That's what we're made of; no other ingredients at all, we're just made of cells... Not a single one of those cells is conscious. Not a single one of those cells knows who you are or cares.–Dan Dennett


●Your cells, in total, contain 249 times more information capacity than all the computers in the world (but redundant)

●“According to [IDC] researchers the total amount of digital information in 2007 reached 281 billion gigabytes.”

●All genes present in each cell (100 ∙ 1012) But turned on and off to make different tissues.

●Your genome is roughly 700 MB

Pre-debauchery #2: Substance

“Meat?”–Terry Bisson

Pre-debauchery #2: Substance● What are we made of ?

● “Solidity” is a hallucination – Feeling EM

● All matter mostly empty space

● “Law” of conservation of mass = False.

Pre-debauchery #2: Substance● “Every atom in your body came from a star that exploded, and

the atoms in your left hand probably came from a different star

than those in your right hand. This is the most poetic thing I

know; you are all stardust. You couldn't be here if stars hadn't

exploded, because the elements, the carbon, nitrogen, oxygen,

iron, all the things that matter for evolution, weren't created at

the beginning of time. They were created in the nuclear

furnaces of stars, and the only way they can get into your

body, is if the stars were kind enough to explode.”

–Lawrence Krauss

Pre-debauchery #2: Substance● Jan Baptist van Helmont – “Fundamentalist” experiment● Conservation of mass almost true.● What he did:

Pre-debauchery #2: Substance● The point: The only fundamental difference between you,

and the air around you is organization.

● Where did this organization come from?

Pre-debauchery #3: Heritage

“Your sweat is salty; I am why.” –Bjӧrk (As the Ocean)

Pre-debauchery #3: Heritage● We're the products of an anti-entropic, invisible

hand process.

● Evolution is category-shattering (ring species)

● Most of you know your parents, grandparents, etc.

● Lots of “greats” → Ocean

Pre-debauchery #3: Heritage“To put it more simply, the second law of thermodynamics is the first law of

psychology: functional order in organisms requires explanation. This high level

of functional organization is not a brute fact of the world, produced randomly or

inexplicably. Instead, this functional organization has a known explanation, an

explanation that is unique, well established, and beautifully principled. Physics

and biology, considered together, inform us that natural selection is the only

known natural process that pushes populations of organisms thermodynamically

uphill into higher degrees of functional order, or counterbalances the otherwise

inevitable increases in disorder that plague ordered systems. In other words, all

complex (i.e., significantly better than random) functional organization in the

designs of organisms traces back to the prior operation of natural selection, and

must necessarily be explained in terms of it.” Leda Cosmides and John Tooby

Pre-debauchery #4: Essence

“Brain cells fire in patterns.” –Steven Pinker, on the Colbert Report

(Asked to explain how the brain works in 5 words)

Pre-debauchery #4: Essence● What does it mean to be human?● Being humans, we're very bad at answering this.● Vitalistic thinking pervades psychology. Clarke's 3rd law.● “Why are some things flammable? Phlogiston!”● Unwrap definition:

“Why are some things flammable? Flammability!”

Pre-debauchery #4: Essence● What does this have to do with psychology?

● “The recency effect, in psychology, is a cognitive bias that results from disproportionate salience of recent stimuli or observations. People tend to recall items that were at the end of a list rather than items that were in the middle of a list.”

● Phlogiston all over again!... But harder to see.

● As psychologists, if we want to answer the question of what it means to be human, we must be careful to avoid circular & vitalistic thinking.

Pre-debauchery #4: Essence● Reductionism: Explaining a phenomenon by recourse to lower

levels of analysis (all else journalism)

● “Good” kind & “bad” kind. Bad kind = strawman.

● Going to reduce humans. Doesn't mean higher levels obsolete.

● Computation is fundamental to what we are.

● “Projectiles, wind-up toys, and computers.”(Non-life, NNS life, Life with a nervous system)

● These categories not remotely crisp.

Pre-debauchery #4: Essence● Metaphor that's not a metaphor

“The human mind can be described as a slow-clockrate modified digital machine with multiple distinguishable parallel processing, all working in salt water.” –Philip Morrison

● Everything we are results from this. Most complicated in known universe.

Pre-debauchery #4: Essence

Essence: We are planet-sized meat-robots, made from organized air which itself was made from stardust. We embody the most advanced computers (or technology of any kind) in the known universe, yet this technology was created by an anti-entropic invisible hand process with no intentional agent required to drive it.

Against this background: What are we?

Pre-debauchery: Summary

Pre-debauchery: Summary

Debauched yet?

“I think it's really tragic when people get serious about stuff. It's such an absurdity to take anything really seriously ... I make an honest attempt not to take anything seriously: I worked that attitude out about the time I was eighteen, I mean, what does it all mean when you get right down to it, what's the story here? Being alive is so weird.”

Good time for a Frank Zappa quote

Language?● They actually do talk, then.

They use words, ideas, concepts?Oh, yes. Except they do it with meat.I thought you just told me they used radio.They do, but what do you think is on the radio? Meat sounds. You know how when you slap or flap meat it makes a noise? They talk by flapping their meat at each other. They can even sing by squirting air through their meat.Omigod. Singing meat. This is altogether too much.–Terry Bisson

● I want to talk today about the particular meat sounds we as humans make: language.

Language● Where did human language come from?● Two main classes of views:

Language is a byproduct (i.e. accidental side-effect)Language is an adaptation (i.e. language was selected for)

● His main idea: Language not directly selected for, but a secondary side-effect, like the whiteness of bones; Bellybutton.

● Stephen Jay Gould → First view

Stephen Jay Gould“I don't doubt for a moment that the brain's enlargement in human evolution had an adaptive basis mediated by selection. But I would be more than mildly surprised if many of the specific things it now can do are the product of direct selection "for" that particular behavior. Once you build a complex machine, it can perform so many unanticipated tasks. Build a computer "for" processing monthly checks at the plant, and it can also perform factor analyses on human skeletal measures, play Rogerian analyst, and whip anyone's ass (or at least tie them perpetually) in tic-tac-toe.”

–From Gould (1979)Panselectionist pitfalls in Parker & Gibson's model of the evolution of intelligence

Gould: Language is a byproduct● Gould regularly accused EP of “panadaptationism.”● True panadaptationism is a straw man.● What would true panadaptationism look like? Carlin joke:

● "Dying must have survival value, or else it wouldn't be part of the biological process."

–George CarlinNapalm and Silly Putty

Gould: Language is a byproduct

● Question: What led to the building of such complexity?

Natural selection made the human brain big, but most of our mental properties and potentials may be spandrels – that is, non-adaptive side consequences of building a device with such structural complexity. –Gould (The Pleasures of Pluralism, pg 11)

● Chess grand masters

● Neural tissue expensive

Gould: Language is a byproduct● Aside: Silver medal for best misunderstanding of EP research● “Detect infidelity and other forms of prevarication.”

(Children and coke example)

● Gold medal: Marshall Sahlins. Described by Dawkins here:

http://bit.ly/goQZPE (6:53 - 9:23)

http://bit.ly/goQZPE

Chomsky: Language is a byproduct

● “For one thing, human language is not particularly or specifically a communication system.” (Interview on nancho.net)

● Undiscovered law of physics● Result of large brains

● Pinker on Chomsky: http://bit.ly/gH7Wwk (1:27-3:33)

http://bit.ly/gH7Wwk

Can language be a byproduct?● Byproduct hypothesis reasonable if and only if language is not a

complicated engineering problem.● This is the famous computer made of tinker toys:

Can language be a byproduct?

Point of Tinkertoy example:

Simply making a computer very

large does not magically lead to

the emergence of sophisticated

computations that solve complex,

adaptively relevant, engineering

problems.

Why am I talking about the Tinkertoy computer?

● To understand this misunderstanding, need a quick digression:● What does it mean for something (e.g. a number) to be

“computable”?

● Logicians have developed formal theories of computation.

● But all the methods they developed turned out to be equivalent!

● Ignore the scary details:μ-recursive computationTuring computabilityAlonzo Church's λ-calculusAbacus computabilityGeneral recursive computation


● This led logicians to the idea of the “Church-Turing Thesis”:“Every effectively calculable function is a computable function.”

● More simply: All forms of computation are “the same”

● Does this imply Gould's computer analogy?


● Pinker and Bloom (1990) rebuttal is important!

Can language be a byproduct?“[Gould's] analogy is somewhat misleading. It is just not true that you can take a computer that processes monthly checks and use it to play Rogerian analyst; someone has to reprogram it first. Language learning is not programming: parents provide their children with sentences of English, not rules of English. We suggest that natural selection was the programmer.The analogy could be modified by imagining some machine equipped with a single program that can learn from examples to calculate monthly checks,

perform factor analyses, and play Rogerian analyst, all without explicit programming. Such a device does not now exist in artificial intelligence and it

is unlikely to exist in biological intelligence.”

–Pinker & Bloom 1990 Natural Language and Natural Selection

● Gould's mistake:

● Two senses in which we can interpret the sentence “The brain is a general purpose computer”

● 1) Neural computation is equivalent to Turing computation (& rest)i.e. Any algorithm run on supercomputer can run on neural tissueAlmost certainly true! Follows from Church-Turing Thesis.

● 2) The brain embodies only one (or a small number) of general-purpose algorithms.Almost certainly false.


● More precise language needed to answer our question!

● When do we need to invoke the concept of adaptation?● Flying fish back to water → Adaptation not necessary (Gravity!)

Flying fish out of water, 50m → Adaptation required (Weird...)

But what is a “complex problem”


● The main question: “Is there a complex, adaptively relevant engineering problem being solved?”

● What makes a theory of adaptation good?

Is AI A? Or I? Both? Neither?● Wait... who decides what is a “complex engineering problem”?

One answer: Artificial Intelligence!

● Another view http://abstrusegoose.com/215

● One view

● Also: http://bit.ly/11eOxA (4:55-5:35)

“It was possible to be an extreme environmentalist only as long as the researcher was not forced to get too specific about how performance was actually achieved. In artificial intelligence, this was no longer possible.” –John Tooby (1985) The Emergence of Evolutionary Psychology

http://abstrusegoose.com/215

http://bit.ly/11eOxA

Is AI A? Or I? Both? Neither?● “The word "natural" is completely meaningless. Everything is

natural; nature includes everything. It's not just trees and flowers, it's everything. A chemical company's toxic waste is completely natural. It's part of nature... it's just not real good food!” –George Carlin

“Artificial” Intelligence● Moravec's paradox in Artificial Intelligence:

[It] is comparatively easy to make computers exhibit adult level performance on intelligence tests or playing checkers, and difficult or impossible to give them the skills of a one-year-old when it comes to perception and mobility. –Hans Moravec

“Artificial” Intelligence● “The main lesson of thirty-five years of AI research is that the hard

problems are easy and the easy problems are hard. The mental

abilities of a four-year-old that we take for granted – recognizing a

face, lifting a pencil, walking across a room, answering a question –

in fact solve some of the hardest engineering problems ever

conceived.... As the new generation of intelligent devices appears, it

will be the stock analysts and petrochemical engineers and parole

board members who are in danger of being replaced by machines.

The gardeners, receptionists, and cooks are secure in their jobs for

decades to come.”

–Steven Pinker, (1994) The Language Instinct

“Artificial” Intelligence● Easy problems:

Chess (Deep Blue & Hydra)Complex mathematics and physics (Calculus, General Relativity)Formal logical reasoning (Propositional, Predicate, Modal, etc.)Diagnosing a disease given a collection of symptoms

● Hard problems: Picking up a glass without crushing itCrossing the street without stepping over a shadow Vision (e.g. Color constancy) Deciding whether a given face is attractiveAny human or non-human abilities traditionally considered “instincts”Language Acquisition and Language Comprehension

● “As a species, we've been blind to the existence of these instincts. Not because we lack them, but precisely because they work so very well.” –Leda Cosmides (2006) “Are we already Transhuman?”

● On Language learnability theory:http://www.mediafire.com/?uhvni93nlnp9hit

● Very easy to misunderstand

● “You keep saying that word. I do not think it means what you think it means.” –Inigo Montoya

● “Learning is a folk psychological concept”

● Does NOT mean learning doesn't exist.

Learning language not so simple!

http://www.mediafire.com/?uhvni93nlnp9hit

● We can now answer this question much more thoroughly.● Whenever we see humans exhibiting behaviors that are:

1) Complexly specialized for solving an adaptive problem

2) Reliably develop in all normal human beings

3) Develop without any conscious effort, and in the absence of

formal instruction (no school required!)

4) Applied without any understanding of their underlying logic

5) Distinct from more general abilities to process information

or behave intelligently

When do we need to invoke the concept of adaptation?

1) Complexly specialized for solving an adaptive problem.

● Hurford (1989) – Arbitrary sign out-competes other strategies

● Nowak & colleagues (1999-2000)Imagine a communication system with one sound for each concept.

● More concepts → More prone to errors. Not “error-correcting”

● Can get around this problem by using finite number of sounds,and string together into sequences – one per concept (i.e. words)

● Nowak et al. also address the claim:

● “Syntax of human language too complicated to confer advantage”

● Imagine a communication system with one word for each message.

● Unbounded memorization problem → Bounded expressive power

● Out-competed by compositional syntax


● Chomsky: “The poverty of the stimulus”● W.V.O. Quine – The Gavagai Problem (1960) Word and Object


● Exposed to this stimulus, you learn that “gavagai” means...

"A rabbit." Right?... or do you learn that it means: "Food""Let's go hunting""There will be a storm tonight" (these natives may be superstitious) "A momentary rabbit-stage""Undetached rabbit-parts""Flopsy" (a particular rabbit)"A mammal""A furry thing""Footprint maker""Habitat for rabbit fleas""Top half of a rabbit"


● Could learn by asking questions:“Is this the same gavagai as that earlier one?”

● Rules out “momentary rabbit stage.”

● How to ask? → Need to learn grammar first

● Okay, how do I do that? → Gavagai problem all over again.

● This is what Quine called “The Scandal of Induction”


● This is what Quine called “The Scandal of Induction”● “Come on, we just pick the obvious one!”


Gotcha! Magical thinking.● “Obvious” only b/c of specialized mental programs that allow for quick

inferences in certain domains. ● If something isn't “obvious” to a simple program...then ?● Hallucinations of what is obvious mesh with each other very nicely● Hard to step sufficiently far out of the matrix to make an objective

evaluation of what is and is not “obvious.”● This is where AI comes to the rescue.

● “The crux of the argument is that complex language is universal because children actually reinvent it, generation after generation – not because they are taught, not because they are generally smart, not because it is useful to them, but because they just can't help it.”Steven Pinker – The Language Instinct

● Some fun examples from The Language Instinct to try to convince you to read it.

2) Reliably develops in all normal human beings

● Pidgin to Creole – Derek Bickerton – Hawaiian Slave plantations● Workers: China, Japan, Korea, Portugal, Philippines, Puerto Rico● No chance to learn each other's languages● Pidgin:● “Me capé buy, me check make”

“He bought my coffee, he made me out a check”or “I bought coffee, I made him out a check”

● “Building–high place–wall pat–time–now time–an' den–a new tempecha eri time show you.”“There's an electric sign high up on the wall of the building which displayed the time and temperature.”


● Children invent grammar not present in parents' speech

● Creole:● “One time when we go home inna night dis ting stay fly up.”

“Once when we went home at night this thing was flying about.”● “People no like t'come fo' go wok.”

“People don't want to have him go to work [for them].”

● Don't be fooled! Not just any old errors.● Case markers, prepositions, auxiliaries, relative pronouns.


● Sign languages – Another creolization

● Spontaneously invented by deaf people

● Deaf infants raised by pidgin-signing parents

● Contains grammatical structure not present in parents signing

● Presents problem for any learning theory that doesn't build a lot in!

● So-called mistakes have an underlying logic, which bring us to:

3) Develop without any conscious effort or formal instruction

● Same phenomenon in grammatical “mistakes” of children:

● Why he is leaving?Nobody don't likes me.I'm gonna full Angela's bucket.Let Daddy hold it hit it

● Grammatical in many of the world's languages.

4) Applied without any understanding of underlying logic

● There are technologically primitive cultures, but no primitive languages

● “I don't got no money.” → “I do not have no money.”“Je n'ai pas d'argent.” → “Je ne ai pas de argent.”

● Perfectly grammatical in French and other languages!


● Inflectional Morphology


● Inflectional MORE-phology


● Inflectional EVEN-MORE-phology


● Inflectional morphology

● Wug → Wugs = Wug + “z”● Splort → Splorts = Splort + “s”● Blorse → Blorses = Blorse + “ez”


● “g” is voiced → “z” is voiced (vocal chords vibrate)● “t” is unvoiced → “s” is unvoiced (vocal chords don't vibrate)● “se” pronounced “s” → morphology adds “ez”


● Who has to consciously remember to do this, each time?

● Broca's aphasiahttp://www.youtube.com/watch?v=1aplTvEQ6ew (0:00-0:30)

● No serious problems with: Motor control, Articulation, Attention, Alertness.

● Syntax & morphology impaired severely – Lexicon spared

● Implies language is not one general capacity!

● What we would expect with very domain-specific sub-components

5) Distinct from other abilities to reason or behave intelligently

http://www.youtube.com/watch?v=1aplTvEQ6ew

● Specific language impairment

● KE family

● “It's a flying finches, they are”“She remembered when she hurts herself the other day”“They boys eat four cookies”“Carol is cry in the church”

● Wug Test“Oh dear, well carry on.”“Wug... wugness, isn't it? No. I see. You want to pair... pair it up.”


● Genetic basis of Specific language impairment

● FOXP2 Gene – Found in 2001 – Chromosome 7.

● Perfectly associated with KE family syndrome and a similar impairment in an unrelated individual.

● Most afflicted family members score in normal range on nonverbal part of IQ tests.


● “Wait... all you've shown is that language can be impaired while other forms of intelligence can be spared.”

● “Maybe language requires more brain-resources than any other task, so knocking out some general mind-stuff could impair language but not other domains.”

● Need Double-Dissociation to rule out Task-Resource Artifact

● Need to find people with: 1) Impaired cognitive ability &2) Spared language ability


Williams Syndrome

http://bit.ly/hpKwT7

(5:46 – 6:29)

http://bit.ly/hpKwT7


● We've got a double dissociation

Whoda thunk?● Other double dissociations suggest the machinery giving rise to

language is functionally specialized beyond our wildest dreams.Possibly most surprising:

● Cubelli (1991) and Kay & Hanley (1994)Double dissociation between vowels and consonants!

● Cubelli's patient CF: Asked to spell Bologna, Tavolino, Fondacaro Ciro

Whoda thunk?● Asked to spell Bologna, Tavolino, Fondacaro Ciro● BOLOGNA

= B L GN● TAVOLINO

= T V L N● FONDACARO CIRO

= F ND C R C R

● Domain-general odds of this? (With simplifying assumptions)● 2-29 = 1 in 536,870,912

Other tests of adaptationist theory of language: Molecular Biology

● Ka/K

s ratio = Nonsynonymous/Synonymous nucleotide substitution

● Imagine a “toy language” where:“A, E, I” each mean “RED”“O, U” each mean “BLUE”Consonants B through M mean “DOG”Consonants N through Z mean “CAT”

● AB = “Red dog” “Red dog” also = AC, AD, AF, ..., EB,...

● Imagine we write a whole page of this boring stuff.● Allow random mutations.


● “A, E, I” each mean “RED” Consonants B through M mean “DOG”“O, U” each mean “BLUE” Consonants N through Z mean “CAT”

● “Nonsynonymous”Some mutations change meaning:The mutation AN → ON has “RED CAT” → “BLUE CAT”

● “Synonymous”Some mutations don't change meaning:The mutation AN → IZ has “RED CAT” → “RED CAT”

● Just like the genetic code: AAA and AAG both specify the amino acid “lysine”


● Correct for base-rate probabilities, can get good idea of whether or not gene was focus of selection!

● Enard et al. (2002) – FOXP2 protein has been target of selection in last 200,000 years! (KE Family's impairment)

● Human FOXP2 protein distinguished from that found in chimpanzees by two amino acid substitutions.

● Only one of these changes unique to humans.● Language polygenic, as we would expect for a complex adaptation

Language as a window into evolutionary psychology

● Language is just one of the things that makes us human.● Philosophers have debated for millennia what it means to be human.● In last few decades, it has become possible to use methods from the

sciences of human nature:

Behavioral Genetics Cognitive (Neuro)ScienceEvolutionary Psychology

to make progress on question where philosophy is “stuck”● “The next round of philosophy may well be written collaboratively

by cognitive neuroscientists informed by evolutionary principles.”–Cosmides and Tooby (1998) “Start with Darwin”

Language as a Window into Evolutionary Psychology

Documents

Transcript of Language as a Window into Evolutionary Psychology