Monday, March 26, 2007

Gagaga and Wawawawa

Disclaimer: I'm talking about pronunciation in this essay, not spelling. So, for the purposes of reading this remember that pronunciation and spelling are not the same thing. Although garbage, for example, is spelled with an 'e' at the end, it does not end in an 'e' for pronunciation. That 'e' is 'silent' - thus for pronunciation purposes, the word ends with the 'dg' sound. Similarly, while the word 'friend' is spelled with two vowels 'ie', it's only pronounced with one. Due to font limitations of the web, I'm not able to give the real sounds - just an approximation.

Gagaga and Wawawawa

First words are a wonderful thing. That is, of course, if you can understand them. The following are words that Tommy used when he was about 20 months. Take a guess at what they could be:

aing
baboe
beh
gagaga
gye-gye
haka
han
kishi
ma-ah
maemae
mih
shishi
wawawawa (my all time favorite!)

Stumped? So were most of our friends and relatives!

aing = orange
babo = backhoe
beh = bread
gagaga = garbage truck
gye-gye = bye-bye
haka = sock
han = hand
kishi = fishie
ma-ah = melon
maemae = banana
mih = mitten
shishi = lotion
wawawawa = helicopter

How on earth did he get so far from the adult pronunciation? Believe it or not, almost all of his pronunciations made sense, linguistically speaking. Tommy was very consistent, applying very regular rules to his pronunciation. The only problem is, by the time he got done applying all the rules, the words themselves bore very little resemblance to their adult counterparts. Our first born is lucky that his mom is a linguist, and was able to decode all things he was doing to words.

Really there were two things going on with his pronunciation. First, he had a set of sounds that he couldn't pronounce – typically late sounds like 's' and 'l' and 'f' were mispronounced. Instead of using those sounds, he substituted other sounds that he could do so 'k' for 'f', 'h' for 's' and so on. Despite these substitutions, the sounds he could pronounce were very typical for a two year old. He could pronounce all of the sounds that 2 year olds can typically use: /b, p, m, n, w and h/. He even had most of the sounds that are often acquired between the ages of 2 and 4 such as /d, t, k, g, "ng"/. True he had a hard time with most of the sounds that are generally acquired after age three – things like /f, y, r, l, s, z, "ch", "dg", v/ (although he did have "sh", which is often a later sound). But difficulties with these sounds is very typical and we're used to that in children's speech. After all we can all understand Tweety Bird's "I tawt I taw a puddy tat!" or Elmer Fudd's "wascally wabbit".

Tommy was so very difficult to understand not because of his sound substitutions but because he altered the sequences of sound, or the 'shape,' of his words. He preferred for his words to end in a vowel, so he omitted the consonants at the ends of syllables and words. Or, if they didn't end in a vowel, he added one ('kishie' is an example of this). He couldn't pronounce two consonants that occurred in a row, so he just deleted one; 'bread' became 'beh' and 'hand' became 'han'. He often didn't pronounce syllables that weren't stressed, especially if they occurred at the beginnings of words. So, those syllalbes just disappeared as the first syllable of 'banana' (maemae) or the second syllable of 'mitten' (mih).

But he did more than that just delete and add. He also preferred for the sounds in his words to 'match', so he'd replace one sound with another from the same word to make all the consonants the same. So, 'backhoe' with a 'b' and a 'k' became 'baboe'. For 'lotion' with it's difficult to pronounce initial 'l', he took 'sh' from the second syllable and used it instead, giving us 'shishi'.

The result of all of this was that his words had a very nice, regular structure for every syllable: Consonant-Vowel (CV), where often (but not always) the same consonants were used in 2 and 3 syllable words (CVCV).

Interestingly, the structure he created for words (CVCV) is one of the most common structures for words across the world's languages. Many language such as Spanish, Italian, Japanese or Hawaiian have very few (if any) words that end in consonants. English and the other Germanic languages (German, Dutch, Swedish, Danish, Norwegian, Icelandic, etc.) are more unusual in that our words tend to be short, but very complex. In Spanish, a 'friend' is amigo – where amigo has a vowel - consonant+vowel - consonant+vowel structure (VCVCV). Compare that with English friend - a short word with two consonants at the beginning and two more at the end (CCVCC)!

What Tommy did with his pronunciation was very typical for a child under 2. He created a template for words (CVCV primarily) that made the words easier to pronounce. This tendency to simplify is so common that a lot 'baby talk' or baby words reflect this tendency for children to simplify both sounds and word structure. Words such as 'wawa' for water, or 'bubba' or 'sissy' for brother and sister are examples of this. And it's probably not just chance that words for 'mother' and 'father' across the world's languages are usually CVCV with the early acquired consonants and vowels: mama, amma, ummah, anaana, dada, papa, ataata, tata, etc. We want our names to be things our kids can pronounce! And thus, even when we have a difficult to pronounce word such as 'mother' (the 'th' sound is learned quite late), we substitute other words such as mommy or momma that our children can pronounce.

So, in terms of pronunciation rules across languages, Tommy's pronunciations weren't all that weird. It was just the sheer volume of pronunciation changes that stymied most listeners. Can't pronounce garbage truck? Well, just delete the 'r's that you can pronounce(so 'garbage truck' would become 'gabage tuck'). Then delete the consonants at the end of 'truck' and 'garbage' since we don't like those either (yielding 'gaba tuh'). Then make all the consonants (and vowels) in the word the same, and voila: Garbage truck has become gagaga! It all makes perfect sense, in its own little world.

Friday, March 23, 2007

Firefighters and Police Officers

"Tommy, you want to play firefighter when we get home?"
"Of course!"

Do you remember a time when firefighter seemed awkward when compared to fireman? My kids don't. They happily play firefighter, police officer, letter carrier and ambulance driver. They've never met a stewardess, only flight attendants.

Does it matter that they're growing up with different words than I did? In some ways, no. Calling someone a police officer rather than a police man doesn't make him less male, if he is a male. But, calling someone a police officer does open up the possibility that they could be female if you haven't met them yet. In this way, the titles we use do matter. I grew up in a world with policemen, firemen and mailmen. I had a paper route and had to collect money from customers. How I hated it when I heard people say "Honey, the paperboy is here!" It denied part of me. Along with the words, there was a tacit understanding that those weren't jobs for girls or women. Only after women started to join the police force or the fire department, did the job titles change. When I had my paper route, I would happily have been a paper carrier, rather than a paperboy or even papergirl.

But, oh how these changes have been decried as they've taken place. Traditionalists have claimed that the new words are grating on the ears, or even absurd. How can we talk about a committee chair convening a meeting, as if an inanimate object could do such a thing! Others comment on the awkwardness of the construction he or she (or s/he) in writing, and note that no one ever says these things. But, the spoken language has a perfectly good alternative: they used for singular if you don’t know the gender. It’s what we’ve done since Old English, and while it’s decried by logicians (a plural pronoun shouldn’t refer to a singular thing), it’s an acceptable alternative.

Furthermore, these arguments miss the point that language changes. It always has and it always will. We no longer use the Old English word gnyrn (sadness, sorrow), we've added they (from Scandinavian), niece (from French) and sheriff (from Arabic). Did the Anglo-Saxons quibble about this new pronoun they or mourn the loss of gnyrn? Did Robin Hood complain that the word sheriff grated on his ears? The person certainly offended him. The word? Probably not.

More insidious are arguments such as those made by self-appointed pundits such as Christina Hoff Summers. She suggests gender neutral language is unnecessary because many languages mark gender. She suggests that to be gender neutral, we'd have to change the whole structure of French, as all French nouns are either masculine or feminine. Or, she argues, man can be specific to males or refer to humans in general, because words can have more than one meaning. But Hoff Summers' arguments are straw arguments. She confuses grammatical gender, used in French, with natural gender, used in English French speakers happily refer to tables as she and desks as he in a way that no English speaker can do with a straight face.

Furthermore, such arguments miss the fact that words retain their multiple meanings, despite the context. While man can have more than one meaning, that doesn't make the male meaning disappear. When we read words with more than one meaning, we call up not just the meaning given by the context, but all the possible meanings. Psycholinguistic experiments have shown that if I tell you that "The diplomat was shocked by the presence of bugs in the hotel room," you will briefly call up both the meaning insect and listening device when hearing 'bug'. Thus, using man when you mean people or humans, will call up the meaning male, even when it's not intended. Why not just say people and avoid the whole issue?

Finally, using fireman or mailman gives subtle indications of our culture and beliefs. Referring to firemen suggests that this is a male job. Girls need not apply. This is no longer the state of the world. My daughter is free now to choose a career other than teacher or nurse. She's already seeing herself in roles that even I, child of the feminist revolution, never could. As she said to her brother the other night: "Tommy, you be the nurse and I'll be the doctor!"

True, some of her dreams are still impossible:

Tommy: "Do you think I could be a firefighter when I grow up?"
Mom: "If you want to."
Maria: "And I'll be a Dalmatian when I grow up!"

While reality may keep her from becoming a Dalmatian, at least her language won't keep her from becoming a firefighter.

Monday, March 12, 2007

Red Light, Green Light



"What would happen if red means "go"?" Tommy asked me at age 3. We were stopped at a red stoplight, and he was pondering the issue. "Good question," I responded. "What would happen? Do you think everyone would run into each other?" "Maybe, if they didn't know that red means go."

While answering his question, what my mind was really thinking was "Aha! Arbitrariness of the sign!" He recognized that the fact that "red" means "stop" and "green" means "go" is an arbitrary convention, agreed upon by members of society. There's no real reason that red has to mean stop. We could indicate "stop" with purple, or orange, or brown. Red could mean "go" or "slow down" or "speed up". We've linked "red" with "stop" through convention. By asking what would happen if this were changed, Tommy was indicating that he recognized this as a convention, or as Ferdinand de Saussure (1857–1913) phrased it he recognized the arbitrariness of the sign.

Depebo, arbre, baum, tree - what do these words have in common? Nothing other than the fact that they mean the same thing in different languages. Each language has chosen its own set of sounds to link to that meaning. Thus the 'sign' (in this case the word) is an arbitrary link to the meaning. This arbitrariness of the sign is a crucial design feature of human language. With a few exceptions (onomatopoeia and sound symbolism), there is no relationship between a word's meaning and its sound. It just is. And that lack of relationship has given us the freedom to link many meanings to sound -- our language need not be grounded in reality. We can talk about ideas, fantasy or even the impossible. The arbitrariness of the sign has also made it possible for many different languages to come into being, each with their own unique links to concept and meaning.

Yet, the concept of arbitrariness of the sign isn't self-evident. 2 year olds, when asked what happens if you call a 'cat' a 'dog' will often respond with 'it'll bark' or something similar, suggesting that they think the cat becomes a dog. For them, word and the thing appear to be one in the same. A 4 year old, however, will say "that'd be silly" or "no one would understand you." They've begun to learn that words and things are separate. A word is just a label. And if it's a label, you can manipulate it. You can change it. You can represent it with sound. Or with letters. Or with a sign. Tommy's recognition that this relationship between red light and the meaning 'stop' is just a convention of society was an indication of growing metalinguistic awareness. Metalinguistic awareness is the ability to recognize the arbitrary and systematic nature of language, to be able to reflect on language and manipulate it, rather than just use the parts to communicate. And metalinguistic awareness turns out to be a crucial component of sophisticated language use – from everything to the ability to make jokes and puns to using metaphorical language to learning to read.

None of this thinking about metalinguistic skills, however, helped me to answer the next questions:

T: "Why does red mean stop?"
M: Well, it just does. It's a rule that everyone agrees upon so we don't run into each other. (Trying to phrase 'arbitrariness of the sign in 3 year old terms).
T: Who made the rule?
M: Umm.. I don't know.
pause
T: I think the stoplight guys made it.

Sounds good to me.

Monday, March 5, 2007

Bricklayer vs. Concrete Mixer

Parents with more than one child are drawn to compare, even if we know we shouldn't. "He's my introvert and she's my extrovert." "She's my mathematician and he's my poet." And we speculate on the source of the differences among our children. "She's got her father's eyes, and he's got mine." "Oh, she's just like her dad!" Learning language is no exception – different children seem to go about the task of learning language in different ways.

Tommy was what I'd like to call a "brick layer" – he started speaking word by word. He amassed a huge vocabulary (250 words or more) before he started to combine words. "Doggy ride" was his first word combination, at age 22 months, while pointing to a picture of Clifford the Big Red dog riding on top of a bus. But, having declared "doggy ride." he didn't combine words again for several more weeks. Just as he was turning two, he moved slowly from one word to two. Then three. He seemed to need to examine and place each word carefully before moving on.

Tommy's bricks (words) were placed long before the mortar came. The mortar, in this case, being the grammatical words and endings that tie things together. So, his first sentences were things like "Doggie ride" "Momma go." Only after he mastered 2-3 word sentences did we get "Doggie rides" and "The doggie rides" came even later. He used names (Tommy, Momma) to refer to people for a long time, "Where Momma going?" "Tommy climb up!" or just plain "climb up!" Apparently, he found pronouns such as you and I unnecessary. He took the big pieces of language and laid them together, brick by brick, then slowly added mortar to build up the language.

His learning of the sound rules of English worked the same way – slow and careful, making sure he'd gotten each part before moving on. When he was very small (20 months), nearly all of his words began with a consonant and ended with a vowel (CV). So, book was [bu], car was [ka] (as was clock), blanket was [baba], bounce was [ba]. Sometimes, stymied by complex words, he resorted to getting the right number of syllables and leaving it at that. So, helicopter was 'wawawawa' and garbage truck was "gagaga".

Given all the changes he made to words, it took a translator to decode a lot of his language. Many of our neighbors and acquaintances were amazed and bemused by my translations. When Tommy announced: 'wu. gagaga. da' and I responded "Oh look (wu), the garbage truck (gagaga) is there (da)," I think they thought I was making things up. But, as time went on, he slowly added consonants to the ends of words. First he added sounds to the ends of words and we got 'gaash' for garbage, and 'tuck' for 'truck.' Then he added sounds to the middle got 'gachidge' for garbage. And finally, well after age 3, he added several sounds to the beginnings of words, and we had 'garchdige twuck' with all the parts of the syllables in more or less a recognizable fashion. As with his sentences, he built his words sound by sound, slowly adding them together. And typical of bricklayers, when he learned a rule, it appeared all at once. So, when he discovered that he'd been missing a whole syllable at the beginning of words, that 'recycling' wasn't just 'cycling', he suddenly began to add 're-' to ALL the words that began with unstressed initial syllables. So, we talked about 'repellers' (propellers) and teacher 'Rechelle' (Michelle). And just for good measure, we added "re-" to some words that didn't need it. So, we talked about 'recycling retrucks' and 'rebatteries'.

Our bricklayer Tommy contrasts strongly with our second child, Maria. Maria is a "concrete mixer". Her language came out all at once, or so it seemed. I can remember her calling up the stairs at 16 months saying "Momma, where are you?" (OK, so it was 'mama wa aaa uuuu?' but the meaning was clear.). She did have individual words, but her progress from one to a several words was much more fluid. She learned both words and phrases all at the same time. I can't tell you exactly when she started doing two word phrases, and it's not just because she's the second child! It's because they appeared nearly at the same time as her words did. So, 'bye plane' and 'that's the eyes' occurred at the same time as book, Abby and no (17 months). Her 'mortar', pronouns such as you or I and the endings on verbs or nouns, appeared quickly too, mixed together with the words. Instead of laying down bricks, she built a whole wall! And we're never quite sure what's going to be in that wall. She looked at a picture of a boy yesterday and say "That boy reminds me of Tommy." Holy cow, what a phrase for a 33 month old.

Unlike her brother, master of the rule, she's much more fluid with her rules. Sometimes she'll add the endings – so sometimes she walks and sometimes she just plain walk, no ending. Sometimes she'll come close to the endings, but not quite – so she 'spelt' her name, sounding quite quaint and old fashioned, while her brother only ever spelled.

With pronunciation, again, her structures came out all at once. No slow building up from consonant-vowel, to consonant-vowel-consonant for her. She took a running leap and jumped into the whole word. Some of her leaps were spectacularly successful. "Garbage" a word which gave her brother fits for over a year, appeared at 22 months as a clear "garbish". On the other hand, "The Laurie Berkner Band" a phrase she attempted at 31 months came out as "wuba babna", a spectacular failure (missing several syllables and substituting lots of sounds) that took us nearly an hour to figure out! And that's the problem with being a concrete mixer – if your wall falls, you don't have any pieces to pick up, you just have to start all over again.

My children, much to my amusement, fit a lot of the generalizations in the literature about variations in style. First borns tend to be 'brick layers' (or referential or analytic learners). They start with words, and build up a large vocabulary before combining words. They apply rules across the board. They resist imitating words. They start by combining content words and add pronouns and grammatical words later. Later borns tend to be 'concrete mixers' (holistic or expressive learners). They start with more phrases (many social phrases), have a more gradual transition from single to many words. They're more variable in their application of rules. Their sentences contain more of the grammatical mortar and pronouns from the beginning.

So, where do these differences come from? Are they real? A good question. Is this another thing to chalk up to first vs. second born? Maybe not. While first borns tend to be bricklayers, girls also tend to be bricklayers (and first born girls even more so). Boys have more of a tendency toward concrete mixing. Gender thus seems to play a role. Or is it how parents react to children – talking differently to boys than girls? Or parenting – do parents of single/first children have more time to label, more time to wait for single phrases to be built up to sentences? Or is it personality? My first born bricklayer is an introvert. He's not going to lay anything out there until he's sure. He's also very interested in things. Labeling things, classifying them is important to him. When he was two, he commented Moon has no doors. Airplanes have doors. My second born concrete mixer, on the other hand, is an extrovert - getting out whole social phrases (where are you? what's your name?) was important to her. She's very interested in people. When she was two, her question about the moon was Where is the moon's families? Did her interest in people and relationships facilitate her language, or is she more interested in people and relationships because she's got the language to communicate about them? The research data is certainly mixed. It's hard to make generalizations, and even harder to find a 'cause' for the variability.

The fascinating thing is that these different approaches do not produce different results in the long run as far as we can tell. Each approach gets a child to language. Everyone has to learn words. Everyone has to learn to pronounce. Everyone has to learn the grammatical bits that hold the sentences together. Whether you acquire these parts as individual bricks and mortar or all in a mix, they do come together in the end. And once the wall is there, it's apparently irrelevant whether it was poured from concrete or built with bricks.