Building a crane to the vocabulary spurt

In the third chapter of Becoming a Word Learner, Linda Smith talks about how children build a crane (her term) to make themselves better word learners.

The notion is that when children learn their first words, they start to notice patterns about those words and create templates for future learning. Her main point is that children seem to fix on shape, rather than some other property, to signal an object's class. For example, chairs—prototypically anyway—have four legs, a seat and a back. This shape cues children in that a CHAIR is a chair. What's interesting is that when researchers cue very young children in on the shape bias by training them, their vocabularies grow faster.

So whatever the exact mechanism may be, children are learning how to learn words by—and this is truly shocking—learning words. Once they get to a certain point, the biases and patterns they've developed seem to take on a life of their own.

How might this relate to learning a second language? I'm not wholly sure, but allow me some speculation. One of the things that foreign language learning materials seem to focus on is inflectional morphology, which makes enough sense. You can't speak the language if you don't know how speakers expect things to be ordered. Latin wants case inflection on nouns. English wants word order. Spanish wants you to be clear about which object you are talking about via definite and indefinite articles. Russian couldn't care. And so on.

But one thing that foreign language materials, so far as I've seen anyway, don't worry too much about is derivational morphology. How do you get from civil to civility? And why can't you go from polite to politity? I'll be reading a paper—and thus blogging about it later—about this subject exactly.  I could be wrong, but I suspect that adult learners are given vocabulary lists that they then create a derivational morphology from. Or at least that's how it felt to me when I was learning Latin all those years ago. Civis became civitas. Aestus became aetstas. Could moralis become mortalitas? And the connection is made, though not without flaws. I think it was then that my grasp on Latin vocabulary started to really firm up from a list of words to memorize to things that behaved in similar way. In other words, I had made a derivational morphology crane for myself. 

Talking toys

Well, it's begun again. In the last week, I've been hearing some odd conversations around the house. My little guy has started talking on behalf of his toys. So I bring up all of that index and symbol stuff just in time for him to start having the words that come out of his mouth stand in for the words coming out of the toy's mouth. Ka-boom! So much for what I had written.

Tonight's conversation?

Little Guy (in the tub): I going potty!
Me (to myself): Oh, no you're not!
Little Guy: I going potty!
Me (getting up to see over the sink): *whew*
Little Guy (holding doll over the boat doing duty as a doll potty)

Icon, check. Index, check. Symbol, um.

So we're reading Becoming a Word Learner for class. And so far it's striking me as an extension of First Language Acquisition—everyone's got a different model.

The first chapter is "Word Learning: Icon, Index or Symbol?", which seems like a good place to start the discussion about learning words. After all, you need to show what it is that people are doing when they learn words. The way the authors, Golinkoff and Hirsh-Pasek, do that is by looking at attempts to teach non-humans human languages. (They say infrahuman, but I don't like the term. It smacks of chain of being, which I detest.) And to get to a human understanding of language, you need to be in possession of what they call symbol. Animals can, for the most part, only manage icon and index. But the part that bothers me is that I can't quite draw the distinction between index and symbol.

An icon of fire

Icon seems pretty straightforward. An icon is a representation of the thing itself. Here's an icon of fire.

See? It's not a fire, but it looks like fire. There seems to be some dispute as to just how much resemblance is necessary, but I'm going to ignore that.

An index of fire

The next remove is index. An index is either something that is correlated with or points to something. So for fire, smoke is an index. Other indexes of fire might be: heat, wood, camping, cooking, matches. So here's a picture of smoke, which is an index of fire.

One symbol for fire
(according to Wiktionary)

The problem comes in with symbol. At this page (a somewhat less in-depth discussion than in G and H-P), symbols are "easily removed from context" and "associated with large sets of other words". Ok, so far so good. I can talk about fire with none being present, as well as knowing that it as an association with other words like smoke, heat, wood, camping, cooking, and matches.

Here's the problem, which strikes me as a father of young children. We talk a lot about the here and now at home, which means that we are talking about things that are not removed from context—particularly with my son (2;5). My daughter has made the leap to things that aren't present, i.e. her upcoming birthday party. So we're kind of defeating the benefit of a symbol. In fact, we're treating words like indexes. We don't say MILK unless there is milk somewhere nearby: or we are trying to get the milk from the fridge into a cup or something very concrete. The other thing is that while we are indexing MILK to milk, we are also indexing it to such things as cups, lunch, cold, cereal, spoons, fridge and the like. So we're somewhat taking advantage of the association with other words, but they too are indexed in the here and now. 

Anyway. What I'm trying to get at is that I'm not seeing a clear line between index and symbol. Maybe at the ends of the index/symbol spectrum of goodness, it's clear. (Oooo, could it be a spectrum relationship?) Maybe as a child's ability to use language apart from the here and now develops, the child develops cognitive ability to make symbols out of indexes.

But there seems to be a lot of messy could-go-this-way or could-go-that-way and begging the question involved with indexes and symbols. If words are symbols, why are they so indexy early on? If words start their mental existence as indexes, then what transforms them into symbols? Do we even need to draw a distinction between index and symbol other than to say that a symbol is an index plus displacement? Or is this just another way that we're trying to separate man from beast without pointing at the actual neurological difference between what humans and, say, language-trained chimps are doing? I don't know. As usual here, more questions than answers. I absolutely promise interesting tools for learning new words in a second language before the end of summer. Cross my heart.

If you're curious, you can browse the book here. Why not? It's pretty interesting so far.

Complementizer development in Latin

As most Latin folks know, oratio obliqua doesn't have a complementizer in classical Latin.

1. Puto    Dalleca a   Doctore     vici.
    I.think Daleks   by the.Doctor to.be.defeated
    I think that the Daleks are defeated by the Doctor.

First, because Doctor Who. Second, there's no complementizer, which is to say that there's no word that corresponds to "that" in English. It's not a big deal. We certainly can skip using the complementizer in English.

2. I think the Daleks are defeated by the Doctor.

Completely grammatical. We might even be able to parallel the Latin syntax and switch to an infinitive.

3. I think the Daleks to be defeated by the Doctor.

Though some native English speakers may feel I'm pushing the bounds of acceptable on that one. Anyway, the reason I bring this up is that at some point Latin picks up a complementizer. It presses quod, which is a conjunction meaning because, into service as the complementizer. So our first sentence might become something like this if it were to show up in the Vulgate.

4. Puto   quod  Dalleca a   Doctore     vicuntur.    I.think that    Daleks   by the.Doctor they.are.defeated
    I think that the Daleks are defeated by the Doctor.

And that's no small matter. Aside from adding the complementizer quod, we made two other shifts. One is invisible, because I couldn't resist the notion of the Daleks in Latin. The noun is neuter so we don't see it shift from accusative in 1 to nominative in 4, so you'll just have to trust me that it works that way. The other thing I did was to shift vici from an non-conjugated infinitive to a standard conjugated verb, vicuntur.

The reason I bring it up is because in my preparation of de Senectute for publication as a reader for intermediate students, I've noticed that quod seems to behave oddly. It's like you could interpret it as either because or that. (And if I think of it I'll dig up an example of this, but not today.)

Lexical acquisition

I start my first (and hopefully only) summer course tomorrow. Lexical acquisition. One of the projects will be a snap. Learn new words in a second language. I doubt that's all there will be, but it seems pretty easy. Find a clutch of words I don't know. If I'm feeling ambitious, which I haven't been lately, I'll pick Persian. If I'm not, I'll pick Greek.

The problem is that the other side of the project is finding new English words. The problem, so far I can see it from before the class starts is that I don't know what words I hear that I don't know. The context of stuff is usually pretty obvious. In reading, I'll have context, but I'll also have that "hm, I don't think I know that one" sensation. The other problem is that if anyone is using unusual words in my life, it's me.

I hope I can blog a lot about the stuff I learn in this class, because it sounds fascinating to me. Thus it will get inflicted on you, dear readers.

Adjective order in Latin

Warning: speculation follows, hopefully not all baseless

As we all know, you can graduate from college magna cum laude or summa cum laude. Latin students can skip ahead, but for those of you who don't know: here's the story. Both of those expressions are prepositional phrases. We'll skip some of the fine points, but here's the outline

1. magna cum laude    great    with praise
    with great praise

Now, you might notice that the adjective is before the preposition, and that's odd by English standards. In ordinary Latin you might get

2. cum laude magna

Yes, the adjective is prone to coming after its antecedent noun, hence the name antecedent. In previous entries, I've mused as to what is going on here. And I don't know that I've got any stunning new insight about why the preposition should move to where it does. In fact, it probably doesn't move to the spec P position. Oh well. Here's an example

3. Cytoriaco radium de               monte      tenebat (Met, 6.132)
    cytorian    shuttle  down.from mountain she.held
    She held a shuttle from the Cytorian mountain (i.e. boxwood).

The adjective Cytoriaco is separated quite a way from its antecedent noun monte, which is behind the preposition—as it should be. And this made me wonder whether there was some rule in how to separate an adjective from its noun. Latin is famous for this sort of word order play in poetry: noun and adjective bracket a line. It's picked up a name: Latin sandwich. Here it is in the Metamorphoses.

4. et    rupit         pictas      caelestia crimina vestes (Met, 6.131)
    and she.broke decorated celestial  crimes   cloths
And she [Minerva] tore apart the tapestry decorated with the crimes of the gods

Just what I want to show you. In 4, the adjective, pictas, comes well ahead its antecedent, vestes. Once again, I've done some casual looking around. The pattern seems to hold. If a noun and its adjunct adjective are separated, the adjective comes first. 

But so what? It is even real? I don't know I'll have to file this under: things to look into.

End of semester wrap up and reflection

And spring semester is over. Or, rather, I went to the last night of classes last night. 

As you may know, I've decided to post my homework on Scribd—at least the stuff that might be of interest to someone. Why? If someone is interested in how language works, my homework might be of interest. So as much as I think I can get away with it, I try to write for you instead of the professor. Whether I've succeeded or not is another matter that I leave to you.

One advantage of posting my homework beyond wider circulation is that I can refer back to it, which I did on a recent paper. I saw something in subject agreement marker syncretism that reminded me of some work with pronouns I had done. And a revelation hit me: I couldn't have done this (easily) pre-internet. I made the connection, referenced it in the paper and had a bit.ly link pointing  back to the previous work. That's really powerful. Building connections between classes, admittedly easier at graduate level, is a major part of education: taking disparate information and synthesizing knowledge.

So what information got poured into my head this semester? Morphology, which may or may not be a thing, and First Language Acquisition. I loved one form the get go and needed some encouragement in the other. Oddly, the classes were taught by husband and wife. They couldn't have been more different for it. 

But now I don't have to tangle with stuff I'm less interested in. Hopefully there will be more time for blogging and crunching through texts. Derivational morphology in Latin anyone?

ho de

My research presentation from today. The professor moderating the symposium wasn't sold on my conclusion, but did like the direction I was taking.

First language acquisition

The research problems here are massive: you've got to deal with small children as your informants. I can't think of too many areas of research where the child—some just a few months old—have exclusive hold on the answers.

As part of the first language acquisition class I'm finishing up, I did a bit of research involving 2-, 3-, 4- and 5-year-old children. There is something very intimidating about realizing that as an adult, you cannot answer the questions. You can ask. You can interpret the responses. But you can't answer. Your subjects may not want to answer, and they're all you've got.

The other wildly difficult thing about first language acquisition research is that a lot of the principles involved are highly intuitive. For example: when applying names to objects, children assume that things have only one name. Researchers call it the Mutual Exclusion principle. Obvious, but you have to put some scientific rigor to it if you're going to use it in your work.

Anyway, what I want to impart is that this area of linguistics is both specialized and accessible for ordinary people.

Here's my final paper for the curious.

First Language Term Paper by sipes23

Theta roles, animacy and passivity in Latin

If this post seems weird, don't sweat it. I'm putting it up so that I can remember it more easily if I need something to research.

Last night we were talking about the difficulties children face in acquiring passives. One of the problems was reversible theta roles, specifically agent and instrument. As we all know, Latin treats these differently in terms of syntax. Which got me to wondering. Does the Latin passive favor animate subjects to inanimate subjects?

My hunch is yes. But in and of itself it isn't a very interesting question. The interesting question to ask is how does Latin passive use relate to later literature in Romance languages? Does passive favor mentioning agents or instruments? Does passive favor animate subjects? How is this similar/different in daughter languages?