in eugene orgeon they call hazelnuts filberts

[tiedyedave]

I am back from a 1.5 week trip to Eugene, Oregon that I didn't tell any of you about, because I'm a bad livejournalist.

I was doing this thing. It was fun. There were 30-35 attendees (not counting the lecturers); they were almost all graduate students in programming languages. It was nice to be in a crowd of people where I could make a joke about continuations and get laughs, as opposed to groans or an unconfortable uncomprehending silence.

There were lots of people from all over the US, and some Europeans as well (Belgium, Netherlands, Germany, France, UK). When Adriaan, James, and John spontaneously burst into an animated conversation in French, I winced at how little of it I actually recognized. I was talking to David later (he speaks 5 languages) about how I felt almost pained by being only an English speaker, when I thought of the perfect metaphor for my situation. To me, speaking only one language feels like being a grown adult still living in my parents' basement. I hadn't realized how deeply uncomfortable I had become with being monolingual until I had that conversation.

So, in my ancient and revered tradition of setting ambitious goals that I will ultimately fail to fulfill, I would like to learn at least two of: Spanish, French, Mandarin Chinese, Japanese, Russian. Not necessarily in that order.

rjmccall, NB: I met a fine fellow named Tom Harke at the summer school, who is a CS grad at Portland State (pdx). I have a feeling (no concrete evidence, just a feeling) that the two of you would get along quite well. If he doesn't recognize "mutual friend Dave", some combination of "summer school", "tie dye", and (especially) "Mao" should do the trick.

Oh, right. During the summer school, I wore only tie dye shirts, bought four new ones (total now 16), and played a lot of Mao. I have this sudden compulsion to make a web page dedicated to my tie dye shirt collection. I also have a much more latent compulsion to start making my own tie dye creations, at long last.

Orc is starting to sizzle and bubble now. Lots and lots of work to be done. We're searching for a new denotational semantics using event structures. I've been thinking a lot about transactions (inspired by Grossman's lectures on AtomCaml) and adaptive workflow (inspired by Hicks's lectures on Proteus), as well as the relationship between Orc sites and diamond types in modal logic. Gobs and gobs of implementation work needed; I need to suppress the urge to reimplement the interpreter in 5 different languages just for giggles, and instead focus on embellishing and stabilizing the existing Java version.

I've been having huge concentration problems, entangled with my usual "cannot reach stable day-to-day life state" difficulties. I hope these issues fade away soon, as the coming months will require a kind of focus typically reserved for long-range astronomical devices and doctoral theses.

Comments

[sleepsong.livejournal.com]

::resists... urge... to knit you a tie die sweater...::

There's supposedly a "sweater curse" that's frequently cited on knitting in which, if you knit your SO a sweater, he/she will break up with you. I wonder what happens when you knit an ex a sweater... ;D

[tiedyedave]

I'm pretty sure breaking up is an idempotent operation.

[sleepsong.livejournal.com]

'K, cool. Want a tie-dyed sweater? :)

[tiedyedave]

Does that mean you tie-dye the yarn and then knit the sweater, or knit the sweater and then tie-dye it?

Also, sweaters are of precious little use to me in Austin, additional sweaters even less so, though I appreciate the sentiment.

[sleepsong.livejournal.com]

I was thinking more of tie-dying the yarn and then knitting the sweater. It'd have a really nice mottled effect. And I'm just looking for an excuse to make another sweater. I like making sweaters. (Not that I'm anywhere near done with my current one, but that's beside the point.)

[yannaboo.livejournal.com]

where do you find all your tie-dyes? I don't see them anywhere but...gift shops, really.

[tiedyedave]

You need to go to hippie towns, like Austin and (moreso) Eugene. Look for incense and arts and crafts, and occasionally tiedye will also co-occur.

[rjmccall.livejournal.com]

My parents lived in Eugene for awhile before I was born — in fact, I believe my brother was born there, although I may be confusing that particular event with West Lynn.

How competently do you intend on failing to learn these languages? Are you fluent enough in French to read, say, papers by INRIA (at least some of which they publish in French)?

[tiedyedave]

I can read well enough to get impressions, but not to actually comprehend. Reading is my strongest point; my listening was always pretty poor - speaking and writing are basically gone now.

Of all the options, French and Spanish would be the easiest, but they would still be a lot of work. The immediate problem with French is that I don't have any useful way to maintain it, whereas Spanish is indispensable given my geographic and demographic proximity to Mexico. In that respect, French is actually the most difficult; Maria is a native Russian speaker, Walter speaks Mandarin almost fluently, and my lil' sis even knows some Japanese.

[platypuslord.livejournal.com]

On the other hand, you are one of remarkably few humans that can fluently speak Orc.

[tiedyedave]

In terms of utility, however, one would probably instead consider something like

(fluency)^3 * (size of speaking population) * (a normalizing constant)

in which case the utility of my Orc fluency is dismal. Being one of the three or four fluent speakers of Orc is not very useful.

I appreciate the sentiment though. :)

[ikeepaleopard]

I'd say $(size of speaking population) should be proportional to n^2.

[tiedyedave]

I was just making a wild guess. I wonder if there are actually models for this sort of thing.

[ikeepaleopard]

I was basing that on this metric, which sounds pretty reasonable to me. And I'm sure there are network effect utility models, but I don't know anything about them.

[rjmccall.livejournal.com]

That metric is for measuring the overall utility of the system, not the individual utility to Dave, which would still be proportional to the number of speakers.

Well, actually, potential utility is wierd because it's not additive; it behaves a lot like probability in that I can derive the same utility from several potential sources. So, Dave might speak to Bar in Orc and learn something, or he might talk to Baz in Orc and learn the same thing, and the potential utility doesn't accumulate. Properly, when deciding whether to engage in some action, I have to derive the possible outcomes, evaluate the utility of each, calculate the probability distribution, and calculate the expected utility — hooray. And of course, human interaction is enormously unpredictable unless you intentionally limit it, which is why Dan's model is somewhat silly.

Hopefully, we can simplify that through assumptions, but that's difficult here:
1. Learning a language may provide intrinsic utility, independent of communication. Orc, for example, allows humans to digest cellulose.
2. Learning a language affects both the subject and the object of one's conversations in unpredictable ways (q.v. human interaction), which in turn affects primarily non-language utility.
3. Learning a language affects whether other people will learn that language.

All in all, modelling utility is hard, let's go shopping.

[platypuslord.livejournal.com]

Your mom.

[platypuslord.livejournal.com]

My model would be:

SUM_people(desired_frequency_of_communication(person)*
        MAX_languages(MIN(fluency(me, language), fluency(person, language))))

This confirms my previously-held belief that English is the only language I need.