Module 6, Language Development

Language

A structured, rule based, system of communication that uses symbols to communicate meaning.

• There are four major components, being that it is symbolic, expresses meaning, is generative , and is structured/rule-based.

Symbolism

• Language uses symbols to represent other things that we're talking about.

  • e.g., “Dog” is a collection of letters that represents and conjures up the mental image of a dog. Similarly, when hearing someone say the sound “dog”, you're immediately picturing a furry animal. American Sign Language is another form of symbolism.

• These symbols are all totally arbitrary.

  • e.g., We all collectively agree upon or have learned that when you hear someone saying the sound dog that they're talking about a dog, but this is all just an arbitrary thing that just happens to be the word that is in our language. If you learned French growing up and you saw the word or heard the word "chien" in French, you'd be thinking of the exact same thing.

  • The sound doesn't matter. The specifics of the symbol don't matter. What matters is that they're representations of this thing that we've all agreed upon as speakers of a given language.

• Therefore, a symbol is not only a written orthographic image, it is anything that represents something else.

Meaning

• Refers to interpreting what someone is saying to you, you're listening to someone communicate through language.

  • You know that they're trying to express something meaningful, and there's going to be lots of things that influence how you interpret that meaning.

  • e.g., If you hear me say, that's my dog and you see me standing next to a golden retriever, you're going to interpret my meaning for my language as saying that I own that animal. If we're at a barbecue and I see you reaching for my dinner plate, and I say, hey, that's my dog. You're going to interpret that entirely differently. You know I'm not talking about a furry animal anymore. I'm talking about a hot dog.

• Context, tone, and other cues all alter how we interpret language.

  • Language expresses meaning and we're going to try and decode what that meaning is not using only the language itself, but also more contextual things to interpret that meaning.

• Pragmatics is our knowledge about how language is typically used in a given context.

  • Our pragmatic understanding of language is something that develops across childhood into adolescence and is a really important part about being a functional language user.

Generativity

• The capacity for language to generate an infinite number of sentences to express an infinite number of ideas.

  • As bizarre as the concept that's being conveyed is, as long as it's using words that I understand, symbols that I understand, and it's presented using the grammatical rules that I know how to decode, I can understand any idea that's being presented to me.

• You could generate an infinite number of sentences that are all unique and different and they're all going to be totally interpretable by somebody else who understands the language.

  • You can both generate and interpret an infinite number of ideas through language.

• e.g., Consider the sentence, “the skunk, wearing a top hat, ate spaghetti.” Despite never having encountered this particular idea or this particular combination of ideas, I'm sure you have absolutely no trouble picturing what this might look like. And I had no problem generating the sentence myself.

• Therefore, we can communicate an infinite number of ideas that are readily understandable by anybody else who is a speaker of our language.

Structured/Rule-Based

• There are implicit grammatical rules that even if you have never taken a grammar class in your entire life, as a speaker of a language, you have a certain implicit understanding of how the grammar in your language works.

  • You know how to use that grammar when you're communicating yourself, and you use that knowledge when interpreting the language of others as well.

  • You may never have taken a grammar class in your life. You may not know the difference between a subject, a verb, an object, or a conjunctive clause. But you know how to use them.

• e.g., Consider the sentences, “Carl was driving the car. Ryan was walking on the street, Ryan got hit.” Now consider this different sentence that has the major actors in the exact same order in the sentence, but communicate something very different, “Carl was hit by Ryan with his car.”.

  • Despite the order of Carl and Ryan occurring in the exact same places in the sentence. This time, Carl was the one walking and Ryan was the one driving the car.

  • What happened here was that the grammatical tense of the sentence changed. You interpreted the subject and the object of the sentence totally differently as well. Therefore you interpreted what was going on dramatically differently.

  • Now you may not know any of the terminology used, but you were able to interpret both of these different sentences differently, and you would have no problem using these grammatical tenses yourself when communicating.

• These are the structure and rules of your language that you know, implicitly, how to use.

• This is unique to human language, and that we don't find in other communication systems and other animals.

Comprehension of Language

Refers to understanding what other people say, sign, or write.

• e.g., If you heard someone out loud say dog, and were able to then recognize that, when you hear that sound, that utterance “dog”, that they're talking about furry four-legged animals, you're exhibiting comprehension.

• When we're talking about someone's receptive vocabulary, we're talking about what words you'd understand if you encountered them. This is what you can understand as the recipient of language.

Production of Language

Actually speaking, signing, or writing to others. Rather than being the recipient of language, you are the producer of language.

• e.g., If a baby wanted to communicate an idea about a dog, or they saw a dog and they just felt like they wanted to say something out loud, and then they themselves would say “dog”.

• This is more than just understanding it when other people use it. This is being able to use it yourself spontaneously.

• We call this the productive vocabulary. These are all the words that you can spontaneously use yourself.

Development of Comprehension vs. Production

• While both these are really important parts of language, they aren't always equivalent. Your receptive vocabulary and your productive vocabulary are not always the same thing.

• Your comprehension skills develop first.

  • Long before babies are able to produce many aspects of language, they are already starting to learn and code and recognize aspects of the language that's being used by people in their surroundings.

  • This is true, even for adults, to an extent, too. The things you'll understand if you're reading, for example, may be dramatically different than what you would actually be able to use yourself in a day to day conversation.

How Generativity Makes Human Language Unique

• e.g., Lots of different species of monkeys have predator-specific alarm calls. When one member of the species perceives a predator, or thinks they see a predator, they produce this alarm call to let the other monkeys in their group know that they need to respond accordingly.

  • These are specific to what predator they think they perceive. They have one call that warns for leopards, where they make this call and all the monkeys will run out onto the ends of the branches, where the leopards wouldn't be able to crawl out to. They have another alarm call for snakes, which when the call is made, the monkeys check the ground and keep alert.

  • While this is an impressive and evolutionarily adaptive form of communication, there is no element of generativity in these alarm calls. A monkey could never generate a complex communication like “Hello, fellow monkeys. There is a green and brown spotted snake on the left branch of the big tree over there by the lake.”

  • There's no way that a monkey could generate any kind of communication that had that level of specificity or that could be very specifically detailed to a given situation. They could not generate an infinite number of ideas to communicate to each other.

  • There's a very specific function for each of these alarm calls, and all of the other monkeys have an instinctive understanding of both how to produce it, and how to respond to it.

• While this is an awesome example of communication, it is not language. Language is specific to human communication. Language is a specific human phenomenon and is not found in other animals.

Phonemes and Morphemes

The basic components in how we describe the sounds and the ideas that people communicate through language.

Phonemes

The elementary units of sound in a language (i.e., the actual sounds that people who are speaking a language and put together to make words).

• The actual phonemes that a language uses may be dramatically different from the phonemes another language may use. There's lots of variability in the basic sounds that language users will use in their language.

Impacts of Phonemes on Language Learning

• Across all the world's languages, there are around 200 different phonemes that humans can use, different sounds that humans can make. But not every language uses the same phonemes, no language uses all of these potential phonemes. Each language actually only uses a particular subset.

  • e.g., In English, there's only around 40-ish basic speech sounds that English speakers make. Those are different from the speech sounds that speakers of another language might need to make.

• This is what leads to speakers of different languages sometimes struggling to learn some of the phonemes or to tell the difference between some of the phonemes that are used in one language, but are not used in their own native language.

  • e.g., In English, we have no trouble, if you're a native English speaker, distinguishing between the B and the P sounds. But if you're not a native English speaker and you're learning English for the first time, depending on what your native language was, it may not be that easy or intuitive for you to hear the difference between these two sounds.

  • This is because at a production level, if you look at what your lips are doing they're really quite similar. They're both bilabial consonants.

Phonological Development

• The process by which people learn about the sounds that are used in their language and how they're all put together is. It is the acquisition of knowledge about the sound system of a language.

• The kind of things you'll learn through your phonological development is what are the sounds that my language uses, what are the important phonemes, and in what order do chains of phonemes typically occur in my language (i.e., phonemes that are very frequently paired together).

• On the other hand, there are phonemes which you will never ever see next to each other.

  • In English, a Z sound followed by a W sound isn't really a thing. This consonant followed by consonant sound is not something you'll encounter in any English words.

  • However, if you're a Polish speaker, there is no problem with this whatsoever. The Polish word for ruthless is “bezwzględny”.

• The bulk of this learning occurs in the very first few years of life. Even before children are able to speak themselves early in infancy, this process of phonological development is already underway.

Morphemes

• The smallest units of meaning in a language.

  • e.g., The word dog features three phonemes, “d-, aw-, and -g”. These are three distinct sounds or three distinct phonemes that I can produce, two consonants and a vowel sound. But individually, none of them are actually in anything. It doesn't convey meaning. All it is at this stage is just sound.

  • But, put it all together and you get a meaningful unit of sound. A meaningful unit of language, which is “dog”. So, a morpheme is the smallest unit of meaning in a language.

• Another really important distinction, though, is that "morpheme" is not just another word for “word”.

  • Words can be made up of one or more morphemes. It all depends on how many meaningful units are found within that word.

  • e.g., You can't break down "dog" into any smaller meaningful units. Now, if you look at the word "doghouse," this word has two morphemes: dog + house. If I said the word dog, or I said the word house, each of those on their own contain meaning.

• One takeaway point is morphemes are about more than just words or syllables. It's about meaningful units of information. Morphemes are about the smallest meaningful units of information.

Suffixes and Prefixes as Morphemes

These are also morphemes because they're also meaningful and they change how you interpret the meaning of a word.

• e.g., “Dogs” actually has two morphemes because “dog” is meaningful on its own. But also that S at the end, you understand as an English speaker, that when you hear an S at the end, it makes it plural. That's a meaningful addition, a meaningful unit of that word.

Semantics

The study of meaning.

Semantic Development

Learning the system of expressing meaning in a language.

• This is like taking what you know about morphemes and the meanings of units and stringing them together in a meaningful way.

• It involves not only word learning, but also how to modify your language, how to express meaning in a meaningful and intelligible way.

Syntax

• The rules for how different kinds of words, such as nouns, verbs, and adjectives, can be combined in a way that will make sense to another person speaking the same language.

  • e.g., If I were to say, "The dog bit the cat," you probably pictured a dog walking over and taking a little munch on a cat. Now, if I flip things around a little bit, I can say, "The cat bit the dog", now you're probably picturing something like the opposite. This time it's the cat that's taken a bite out of the dog.

  • It's not just the order of the words that we're using that dictates who delivers the action and who receives the action. I could just as easily say, "The cat was bitten by the dog". Here it's the exact same order of cat and dog as in the previous example. But this time, it is, again, the dog who would be biting the cat.

• The key here is that you understood all this because you have an understanding of how English grammar works and how to interpret the meaning of a sentence based on the grammar being used.

  • It's not just about word order. It's about grammatical tenses.

  • The really fascinating thing here is that you probably never received any formal training in grammar. You may have received some, but you wouldn't need to. This is because the syntactical development of humans is really an intuitive and fascinating process.

Syntactical Development

• The process by which humans learn to understand and later follow the syntactical rules of their language. They come to learn and understand the grammatical rules of their language.

• The key here is that this is intuitive and effortless.

  • From the time babies are very early in infancy, they start picking up on some of the syntactical ordering and rules that are emerging in their language. And as they proceed across development, they intuitively and effortlessly learn how to understand and later use these syntactical rules.

  • It doesn't require a teacher directly instructing them how to use the past tense or how to use S to make something plural. Children will just intuitively and effortlessly absorb this information and learn how to use it as part of their syntactical development.

Theory of Universal Grammar

• An explanation for the intuitive and effortless nature of syntactical development, proposed by Noam Chomsky.

  • This contends that there's a set of highly abstract unconscious rules that are common to all human languages. So humans evolved to have this kind of unconscious understanding of how to structure their language in a grammatical way, such that you can communicate in different tenses, in different word orderings while still maintaining certain meetings, etc.

• Chomsky came up with this idea of universal grammar and he put forward that while there are superficial differences between languages, at a fundamental level all languages have the same innate concepts baked into them. How they arrange themselves at the surface level might be a little different, but they all have them.

  • They all have subjects, verbs, objects, tenses, etc.

• Chomsky argued this must mean that these components of language and this understanding of how to use language has evolved to be a central part of humans innate communicative abilities.

• Chomsky also came up with a funny example of how humans can understand grammar even when the content of a sentence doesn't actually make that much sense.

  • He famously said, "colorless green ideas sleep furiously."

  • Does this sentence make sense? Like, does this mean anything? Is there anything about this idea that you can actually visualize? The answer is no. You can't have green that is also colorless. Ideas don't sleep. And I don't know if anything sleeps furiously.

  • But if I were to ask you, does this sentence grammatically make sense? You'd probably say yes. The order of nouns, verbs, adjectives, and adverbs all make sense perfectly. The conjugation of the tense is perfect.

  • The interesting case here is that we have a case of a sentence where grammatically it makes sense, but the content doesn't make sense.

Example of the Theory of Universal Grammar

• Another interesting case of this in popular culture is Deepak Chopra. He's sold a lot of books and is pretty prolific as a writer and a speaker.

  • He's also received a lot of criticism from the scientific community who accuse him of misusing scientific terminology. And basically just saying things that don't make a whole lot of sense.

• One of these is an actual quote by Deepak Chopra. And the other two are not.

  • “Attention and intention are the mechanics of manifestation.”

  • “Each of us inspires immortal actions.”

  • “Self power is mirrored in spontaneous abstract beauty.”

• The real Deepak Chopra quote is the first one. The other two were actually generated randomly using a bunch of words that Deepak Chopra has known to have used and strung together in a grammatically sensible way, even though they don't actually mean anything.

• A computer pulled these words out of the blue and just structured them in such a way that they grammatically make sense, even though there's actually nothing at all meaningful about them.

How We Decode and Understand Language

• In a paper by researchers right here at the University of Waterloo called On the Reception and Detection of Pseudo-profound Bullshit, they looked at individual differences that were associated with whether or not people attributed meaning or intellectual depth to these nonsensical "bullshit" sentences.

• This all comes back to the idea of generativity in language. Language has the generativity property, which means that it can generate an infinite number of sentences and express an infinite number of ideas. And critically, because of the underlying syntax, because of human understanding of how to structure a sentence grammatically, we can actually generate sentences that don't actually mean anything.

“Language” in Non-Human Animals

• People talk about language in lots of different kinds of species.

  • e.g., People talk about dogs' language because dogs can listen to humans instructions, can learn to respond to certain commands so some people might say that dogs have their own language or that dogs understand human language. Dolphins have really sophisticated means of communication. Ants as well seem to work in really coordinated ways and seem to be able to share information with one another which might lead some people to say that they have their own language. Monkeys seem to have really specific alarm calls. Some people may argue that apes have learned sign language and other aspects of what would appear to be language. If you ever wondered why the smell of freshly mowed grass is such a nice fragrant smell, this is because this is a communication on the grasses part.

• These are all really interesting forms of both the understanding and in some cases, the production of communication. But is it language? No, it is not language.

  • Because language is very specific to humans. So to answer the question, who learns language humans? Humans do.

  • We do not have the evidence to support real, deep, full-fledged language learning in any other species other than humans.

• People that liken these forms of animal communication to language are downplaying or missing out on some of the important nuances about what makes language in humans so remarkable. There's so much mind-blowing complexity to the story of how humans have language that by likening any kind of communication in animals to language, we're diminishing just how fascinating language really is.

Examples of Something Similar to Language Learning in Animals

• Kanzi is a bonobo. A bonobo was once classified as a subtype of chimpanzee, but is now recognized as being its own unique species. And humans, chimpanzees and bonobos all share a common ancestor.

• Kanzi was a really remarkable case in the world of the study of language in animals. He learned to use lexigrams as researchers attempted to teach his mother, Matata, to use the lexigram system. While they were actively trying to teach Matata, Kanzi as a little baby was just sitting in her arms.

• A lexigram system is a series of pictographic symbols, and these are equivalent to words. They're symbols that can be used to represent real things.

• While Matata never really picked up the lexigram system very effectively, Kanzi took to it really well.

  • He'd be able to find the corresponding lexigram that matches that object. And he's estimated to have a symbolic vocabulary of over 300 words.

Caveats to Kanzi’s Example

• Important note however, Kanzi actually uses far fewer symbols spontaneously or on a regular basis.

  • When you put him to the test, he has over 300 words in his vocabulary, it's not like he's using this vocabulary the same way a human does.

  • A human will spontaneously babble about literally any given topic. If you look at children early in their language development, children talk all the time.

  • Kanzi does not do so as much. He uses a much smaller subset of words to make simple requests. But he does not actively use his full symbolic vocabulary on a regular basis.

• Not to slag on Kanzi too much, he actually can follow verbal instructions. There's some really cool examples or tests that are showing that Kanzi cannot just use his board, but can actually put the human symbolic words into use (e.g., telling him to put the soap on the ball”).

• What's even cooler is that this happens even when the person talking to Kanzi is wearing a welder mask, so that Kanzi couldn't be looking at their facial expressions nor following their long line of sight.

  • The only thing Kanzi has to rely on here is the words that he's hearing and still he's able to follow verbal instructions.

• He seems to understand verbal symbols, spoken words in addition to his lexigrams. However, his understanding of syntax, which is that really cool aspect of understanding grammar and how to combine symbols into a sentence that humans have. His understanding of syntax is pretty unclear and all signs point towards it being pretty limited.

• While Kanzi is a phenomenally cool case of learning and cognition in a non-human animal, what he's doing here is not human language.

  • He's missing out on a lot of the really core characteristics that we've already talked about (e.g., generativity, syntax, etc.).

  • These are those unique things about human language that are not found in the communication systems of other animals.

• As of right now, we don't have strong evidence that supports language in anything other than humans.

Language Development in Humans

Language is a human specific phenomenon. It's found across the globe in all human cultures and societies. If you're a human, you will grow up learning language, if you're typically developing.

Sensitive Periods for Language

• The window of time during which language develops readily and relatively effortlessly in children.

  • It ranges from birth up until somewhere in elementary school. The exact end of this window is a little bit fuzzy, but it is somewhere around the ages of seven or eight.

• During this sensitive period, children can not only learn one language, but they can learn multiple languages in parallel without any palpable drawbacks.

  • If a child is growing up in a bilingual or a trilingual household, children can very easily learn all three languages, in parallel, at the exact same time with very minimal drawbacks or confusions.

• A study looking at adults living in the United States who were originally from either Korea or China and then immigrated to the United States at some point, either in childhood or early in their adulthood. The question here was, how does when you moved to the United States and started learning English impact your adult mastery of the English language?

  • Between the children who were born in the United States and the children who immigrated to the United States between the ages of three and seven, there's actually no difference in their overall level of mastery. Despite only arriving in the United States years later than the children who were born there, the children who arrived between the ages of three and seven still mastered the English language to the same level of proficiency as children who were born in the United States.

  • The reason for this is because they started learning this language during their sensitive period for language development.

  • We would refer to their mastery of English, or we would refer to English as being one of their L1 languages. This is one of their native languages or a language that they understand so implicitly because they learned it at such a young age, that it is just an integral part of their understanding of language.

• After this sensitive period for language, language acquisition doesn't really work the same way. It doesn't happen as effortlessly or as fluidly.

  • If we look at the level of mastery for children who arrived, or for adults who had arrived in the United States when they were between eight to 10, 11 to 15 or seven to 39, there were late arrivals in the United States, their mastery of English is much lower on average than children who started learning at a much earlier point in development.

  • This is because they started learning English outside of their sensitive period for language development.

  • For these people, we would describe English as being their second language. If you develop the language outside of the sensitive period for language, we would describe it as your second language and not your native language.

Infant-Directed Speech / Motherese

Another interesting phenomenon in how adults talk to babies, referring to a warm and positive way of speaking that really highlights positive emotions.

• It's typically higher in pitch and has much slower enunciation, often swooping between high and low pitches. It's also accompanied by exaggerated facial expressions. So lots of big smiles, big open eyes and exaggerated mouth movements.

• It's exhibited across many, but not all human cultures.

  • The reason for this is it may just be a really adaptive and effective way of communicating language to babies. It draws infants' attention to the speech by highlighting the contrasts between the speech sounds or the phonemes, so by dragging out and swooping up and down, you are kind of highlighting what are the important contrasts between the different sounds.

• It turns out that this improves infants' word recognition.

Infant-Directed Speech / Motherese Across Cultures

Not all cultures in the world use typically infant-directed speech. Even within a culture, some families may not particularly speak to babies using infant-directed speech.

• You may be asking, well, what about the kids who don't receive infant-directed speech, do they still learn language? The answer is, yeah, of course.

• So long as a typically developing child receives that input in their life, as long as there are people speaking language around them, children take to language easily.

Speech Perception

Before an infant can start producing language themselves, they need to be able to break down the speech that they're hearing, the continuous streams of speech that they're hearing all the time, into its important pieces, so they can learn those pieces and learn the important sounds of the language.

Categorical Perception

• Humans perceive speech sounds as belonging to different categories.

• Phonemes are not interchangeable and they are distinguishable. They're always different from each other. But some sounds are bilabial sounds, meaning they involve similar movements in the lips.

• The major distinction happens when you actually start making a sound with your larynx. The distinction here is in the voice onset time, or the time between air passing the lips and the vocal cords vibrating.

  • It turns out that humans draw this distinction between which of the phonemes they're hearing based on the 25 millisecond mark in vocal or voice onset time.

  • e.g., If the voice onset time of that bilabial phoneme is shorter than 25 milliseconds, then we interpret it as a “B” sound. If that voice onset time is longer than 25 milliseconds, we categorize it as “P” sound. When you get down to exactly 25 milliseconds, people aren't really sure how to interpret it.

Example of Categorical Perception

• A study by Imus et al. hooked babies up with these really cool pacifiers for measuring the rate and then the intensity of their sucking on the pacifier. And they used this as a measure of the children's excitement or heightened attention.

  • They would present them with sounds, long repetitions of sounds. They'd start them off with something that was clearly a P phoneme. At first kids would be pretty excited. They'd start sucking pretty happily as they listen to this new sound, but eventually they would habituate to this sound.

  • What happens then is they would present them a different sound. And so as adults, we would interpret this now as a B sound based on its voice onset time.

  • And the question was, would infants recognize this as a change in the phoneme that they were hearing? If so, you'd expect a big jump in their sucking behavior.

  • Turns out yes, they do. So just like adults, the children dis-habituated to this new stimulus, indicating that they saw this as being a categorically different sound to what they were hearing previously.

• In another condition, they made a similar change in the VOT of the sound the infants were hearing, but this time they went in the other direction. It was the exact same change in the VOT stimulus, only now it's not crossing that 25 millisecond threshold. As an adult, they still categorize this sound as a P sound.

  • The question was, do infants similarly categorize this still as a P sound? And the answer is yes, they did. They did not dis-habituate to this new sound, which tells us they still heard this as a P sound, despite the fact that it had this change in its VOT.

  • What we learned from these two conditions is that just like adults, that 25 millisecond mark is the threshold between distinguishing the P sound and the B sound.

  • These are one and four month old infants, these are really fresh shortly after birth babies, and they're already drawing these categorical distinctions between phonemes.

Categorical Perception in Infants vs. Adults

What's maybe even more fascinating is not only do they draw categorical distinctions between phonemes, it turns out they're actually better at distinguishing between phonemes than adults are.

• Rooker and Leland created a study that uses the conditioned head turn paradigm. You condition babies to think that, or to learn that whenever they detect a change in a stimulus that they're presented with, so the same stimulus is presented for a while. And as soon as it changes, you present them with a really cool other stimulus off to the side.

• In this case, it was a bright light that would go off. A nice little musical tune would play and a happy little Teddy bear would do a cute little dance. You condition the babies to turn their head when they detect that change in stimulus, in anticipation of the dancing Teddy bear.

• Rooker and Leland presented babies with a first along a series of a particular phoneme in English like “D”. The babies would eventually habituate to this sound and then you'd change things up and present a different English phoneme. This is “DG” like in the word sludge. So this is similar to D but is in English an important distinction.

• As adults, we would identify that these are two different sounds. And sure enough, when 12 month olds are presented with this change in stimulus and another English phoneme, that's really important, they turn their head.

Categorical Perception in Sensitive Windows

• Importantly, in another condition, they did the exact same thing in the beginning, but then they played a different phoneme name and importantly, this phoneme is not an important English phoneme.

  • The question is, would 12 month olds turn their heads after detecting a difference here? Or would they just assume this is a D like any other D sound? It turns out that's exactly what happens.

  • By 12 months of age kids do not turn their head because they think “this sound is just like a D, there's no difference here. So the Teddy bear isn't going to dance for me”. They did not notice the difference.

• Now they did it again. The exact same procedure, but this time with eight month old infants. So four months younger than the children we were just talking about.

  • These infants, just like all the infants in the study, were raised in an English speaking environment. This would not be any English... Or this would not be a phoneme found in the language in which they'd been immersed in their entire lives.

  • But nonetheless, these eight month olds, unlike the 12 month olds, did turn their head, suggesting that they did hear a difference between this new non-English phoneme and the D sound they were habituated to.

  • In this sense, eight month olds are actually better than 12 month olds and adults at noticing the difference between these phonemes.

  • This is due to something we've talked about in other modules, which is perceptual narrowing. So this is the phenomenon whereby as children develop, they become more finely tuned to the important categories in their language.

  • But as like a trade off, in doing so they lose the ability to detect differences other than those important contrasts in their language. At a certain point in your development your brain has automatically learned the really important phonemic distinctions in your language. Iit starts blurring between everything other than those important distinctions.

  • But at eight months of age, you haven't reached that point yet. And you're still just able to notice any of these phonemic distinctions between different phonemes from all of the world's languages.

• This is one of those things that relates back to the idea of sensitive windows in development, or sensitive periods for language learning whereby the language you're exposed to in this first year of life is going to influence how you process speech sounds for the rest of your life as well because during this window you're really, really sensitive to the sounds that are appearing in your surroundings and as you develop in your surroundings, you learn to hone in on the important ones and disregard the rest.

Word Segmentation

• Refers to breaking things up into words. The real meaningful units that they can attach to concepts and verbs.

• The problem with word segmentation is that while words are really distinct when they're written down, they actually kind of run together when they're spoken out loud.

  • The pauses, if there are any, the micro pauses in my speech, are as likely to occur in between syllables within a word as they are between two different words.

  • It's not like a baby can just attend to spaces to try and figure out where one word ends and the other word starts.

• The question becomes what strategies or how do infants identify individual words from these contiguous strings of words. If they're going to be learning the names for objects in their surroundings, for example, how are they going to pull that apart from the sentences in which those words are embedded?

Prosodic Cues

• Prosody is the characteristic rhythm and intonational patterns in a spoken language.

• There are certain inflections to the words when you're speaking that are common to a particular language. Another word to put this is the stress patterns. Where do you put the stress on different words or different syllables within a word and within a sentence?

  • In English, it's often at the beginning of the words. You could use this as kind of a tool for identifying the start of a new word. In Canadian French, for example, it's the opposite. It might be at the end.

• You can tell that these prosodic cues are meaningful and also specific to the language that the baby is learning.

Distributional Properties

• Another solution is babies may attend to the distributional properties of the speech sounds in their language.

• Certain sounds are more likely to occur together within a given language.

  • We've looked at how babies use statistical learning with very short presentation times to learn that certain syllables occur frequently next to each other, whereas other syllables do not frequently occur next to each other.

  • Babies can use this information about these infrequent pairing, and say “that was kind of a weird two syllables to hear next to each other, maybe that's because those are two different words”. So they use these infrequent pairings as a cue for the transition from one word to another word.

Contextual Cues

• Babies can also use contextual cues and other known words to direct their attention.

• It's basically a process of elimination a little bit. Or in otherwise put orienting your attention to certain words that you know and using that to interpret the other words in the sentence.

• For babies, there's one really important context cue that babies use, which is of course, their own name. One of the first things a baby is going to learn to meaningfully orient around is their own name.

  • e.g., Imagine here we have baby Jerry. Jerry is looking at this new object that currently he doesn't know the word for, but he frequently hears his mom saying “it’s Jerry’s cup”.

  • Jerry is going to early in his development learn that he is Jerry or that Jerry is the word that corresponds to him. And he'll use that to kind of quickly learn that if it's paired with him, whatever object he's interacting with at the time, that is probably the word that corresponds to that object.

  • And if it were some other name that Jerry was less familiar with or didn't have as much salience for him, such as John, he likely would not learn the word so quickly.

Precursors to Speech Production

• Before children are able to speak in words or sentences, everything starts with simple vocalizations and sounds.

• Right after birth, the only sound babies can really make and the really only important one they would need to make is crying.

• Within six to eight weeks this kind of flourishes into a bunch of different non-linguistic, but still communicative signals like cooing, grunting or raspberries.

• Eventually what's really interesting, after children have had some exposure to the phonemes that are used in their language, they begin babbling. Babbling is the repetitive production of speech.

  • If you were like a six or seven month old, and you notice them making noises saying “ba ba ba ba ba”, these repetitive consonant vowel sequences, this is actually a really important development in their language development because they are practicing the production of important sounds they will use in their language.

• For babies using sign language, even signing occurs early in language development, at six to 10 months of age.

  • Even children who are not using a spoken language, but are learning sign language, will babble as well.

• There are other signs of growing communication skills before kids start speaking.

  • This includes using more advanced gestures and pointing.

  • Their joint attention and their gaze following gets a lot better.

• They're a lot better at cuing into what other people are talking about or paying attention to and turn taking. You'll see children who are engaging in back and forth with their parents, even if they're just babbling and not using words just yet, they're starting to learn that when you're interacting with someone, you take turns.

Development of First Words

• Out of all the things they see in their surroundings and out of all the things they hear other people using, they need to figure out what words map onto which things in the world, starting from scratch

• Word learning begins in the first year. In the first year of life, infants are like little information sponges, and this is particularly true for language. They're sucking up lots of information and their perceptive vocabulary is going to be growing all the time. They're sucking up lots of information and their perceptive vocabulary is going to be growing all the time.

• However, first word production really begins leading up to the first year between 10 months and 15 months with lots and lots of variability between kids.

Holophrastic Period

• This is a window after children have learned their first words, in which children use single words to communicate whole ideas. It's like a whole phrase bundled up in one single word.

  • e.g., A kid might just say something like, "drink" to indicate, "I'm thirsty. Can I have some juice?".

  • It's using a single word to communicate a whole idea during this period when they don't actually have all that many words at their disposal to use.

• The phenomenon of over extension is when babies have a newly learned word and they really want to use it, but they end up using it a little more generally than they ought to.

  • e.g., A baby might believe that juice applies to literally any beverage, even like a Cola. Or they might think that any furry animal can be referred to as a kitty.

• On the other side, there's an under-extension, which is using a word in a more limited way than you ought to.

  • e.g., A child might learn that kitty, or might believe that kitty only applies to their household cat, and they wouldn't know that Kitty is applied to any other cats you see in the world as well.

Vocabulary Development

• After learning their first words, babies will slowly accumulate new words into their vocabulary until the point of around 18 months, where toddlers will have an average of about 50 words in their vocabulary.

  • Other kids may not speak until around the age of 18 months or after. There's lots of variability between kids.

• On average, again, from there, the word learning increases really rapidly.

• Our first words occur on average around about 14 months.

  • Then we have the vocabulary spurt. This is this window beginning 13, 14, 15 months, and extending until around about 25, the average of being around 18. This is a period when children start to learn words at a much more rapid rate.

  • This was originally proposed to be this point in development where suddenly something clicks into place and kids have this period of exponential growth that happens all of a sudden. This was proposed to be an important moment in the child's development when suddenly something clicks into place cognitively, and they're suddenly able to learn how to map words onto reference totally effortlessly and at an exponential rate.

  • More recently, the specifics of what happens here, this change, has come into question a little bit. So recent reviews suggest that it's really not all that universal.

  • There's lots of overlap between when some kids would be having their vocabulary spurt and other kids wouldn't have even been speaking yet. So it is odd to talk about this as being this universal developmental tipping point.

  • At the same time, it's also not all that exponential. It may have been a bit of an over exaggeration to use the word exponential, to assume that it's suddenly this explosion of words.

• While it might be a little bit of an exaggeration to call this an explosion or an exponential growth centered around a single point in their cognitive development, it is still totally safe to say that children exhibit spectacular feats of rapid word learning in their second year of life, starting around the 18 month of age on average.

Word Learning and Building Sentences

• The question really becomes how do children learn new words so quickly at such a young age? How are they able to do that at such an early point in their development?

• Learning all the words that you'll need to get by in your language is no small feat.

• What’s most remarkable about the vocabulary spur period is just how quickly babies are able to learn words just from minimal exposure to the pairing between the word and the thing it refers to.

Fast Mapping

• Rapidly learning a new word simply from hearing the contrastive use of a familiar word and an unfamiliar word. This kind of relies on this idea of mutual exclusivity.

• Monolingual babies at least have the belief that every object has one word for it. If you know that word and you hear an unfamiliar word. That unfamiliar word has to map onto something, you don't already have a word for.

  • e.g., If you present babies with two objects, one of which they know the name of like their cup and something weird they've never seen before. If they hear you using the word spatula around them, they can be really quick in mapping on this new word, spatula to this object that they see.

• One of the really interesting things is that amongst babies raised in bilingual households, they don't seem to as easily have this mutual exclusivity rule because they're learning two languages in parallel. They understand that an object can have multiple different names, in different languages.

  • They understand that different people will call different objects, different things at different times in different languages.

  • Obviously that's not a big impediment for bilingual babies because they still go about learning, not just one, but two languages very quickly and very effortlessly, and that's probably by using many of the other phenomena.

Cross-Situational Word Learning

• Another process by which children learn new words, and refers to using repeat co-occurrence across situations to map new words to the reference.

  • What the cross situational word learning describes is how over time babies kind of build up increasing amounts of evidence that supports a word mapping onto a particular thing.

  • e.g., If a baby sees, for example, these two things that they don't have words for yet, but they hear someone say, "look at dax" And then over time they have repeated experiences where they keep hearing the word dax used when there's a particular shape in their surroundings. Well, over time, children will increasingly build up evidence that supports the conclusion that this thing is probably called a dax.

• This is just another example of statistical learning. Something that we can see is really, really important for children's cognitive development and particularly their language development.

Pragmatic Cues

Involve using social information to learn new words. This a really useful way of learning new words is to pay attention to what other people in your surroundings are doing.

• One cue that children might use is eye gaze.

  • e.g., Imagine you have a child interacting with an object they've never seen before. There's also a box that appears to be covering something. The experimenter comes over, takes a peek under the box before the children have a chance to look at it and says, "look a dax." Then later on the object within the box is revealed.

  • What we found is that children, as young as 18 months are likely to learn that object, that they didn't see, that the person looking at beneath the box is a dax. Even though they don't know the name for the other object, based on the experimenter's eye gaze when they were using the word, dax, children are able to draw that connection and label the new object a dax.

• Children can also use expressions of emotion as a pragmatic cue.

  • e.g., An experimenter could tell a child, "I'm looking for a dax." Before peeking under two boxes. After peeking under the first box, they show a really sad face, but don't use any language to describe what they saw. Then they peek under the second box and similarly don't use language, but show a big happy face afterwards, the contents of the boxes are then revealed and 18 month olds are more likely to label the second object a dax.

Syntax

The grammatical structure of one's language.

Syntactic Bootstrapping

• Using that grammatical structure of whole sentences to figure out the meaning of new words.

• In one study, they looked at that or rather they presented infants with a video that featured two slightly horrifying characters behaving in kind of interesting unique ways.

  • As the child was watching the video, the experimenter would read out a word to them, like, "the duck is crowding the bunny." The interesting thing here is that in this video, both the duck and the bunny are doing different things. The bunny's kind of stretching out on the floor while the duck is kind of pushing their head down from behind.

  • What children were able to do here is look at the structure or the syntax of the sentence, the duck, the subject is crowding the verb, the bunny, the object, and interpret that.

  • “Crading” must be the thing that the duck is doing, not the thing that the bunny is doing. If the duck is crowding the bunny, crowding must mean pushing somebody down.

• Using something as subtle and as implicit as their understanding of the grammatical structure or their emerging understanding of the grammatical structure of their language, children as young as two years old, are able to interpret this and quickly learn a brand new verb.

Children’s First Sentences (Telegraphic Speech)

• Now they're moving beyond simply learning words and starting to string them together into some early sentences that connect multiple words.

• When children are first forming beyond their one word sentences, moving beyond the holophrastic period, children engage in what's known as telegraphic speech.

• These are early sentences that are often only two words featuring only the core elements of the intended communication.

• It's called telegraphic speech, because it's named after the Telegraph where people would have to pay by the word to send a message using a telegram. As a result, they would usually rather than having a full sentence with multiple words, they would condense things down to only the core words and necessary words, just like kids are doing when they're forming these early sentences.

  • e.g., “Eat cookie” or “Brody sleep” or “daddy sit”.

• The really interesting thing about telegraphic speech is that the order in which children present their words in these simple two word sentences is not arbitrary. Rather it reflects their emerging understanding of syntax.

  • If they were able to form a full sentence, they might say, "I would like to eat a cookie or I'm going to eat a cookie." Here the child is saying, "Eat cookies." Rather than, "Cookies eat." They're piecing together the key words in the right order.

• It's only later when their language develops a little more sophisticated that they'll be able to connect those words a little more smoothly into a fully fledged adult sentence, but still the order of their words, it reflects their emerging knowledge of the syntax of their language.

Regularization

• Another phenomenon we can observe in children and as they're forming, their first sentences are over regularization. This is when they use their newly learned grammatical rules or grammatical modifiers in new or irregular circumstances where they aren't entirely correct.

• What we mean by this is that languages are often complex (e.g., there are lots of irregular conjugations of verbs, words don't follow the same rules for pluralization that other words do), but kids are making mistakes that still reflect that they're learning these general rules and there's misapplying these general rules in these exceptional circumstances.

  • e.g., Let's imagine a kid saw these two fellows crossing the street. So moose one and moose two, how might they describe this? They might say, "Look, mooses." Now as an adult English speaker, you know that the plural of moose is for some confusing reason, moose.

  • However, the fact that the child is making this mistake isn't silly or dumb. This is actually them using very apply that rule that they've learned applies when pluralizing moose downs. And they're just applying it to this case that just so happens to be not how we do it as adults.

• We can't really knock them too much for that because they're applying the general rule correctly. It's just that this is an exceptional circumstance.

  • This is an example of over regularization. Over or misapplying these general rules that are typically correct, but are not correct in all circumstances.