“At any given moment, our most complicated machine will be taken as a model of human intelligence, and whatever media kids favor will be identified as the cause of our stupidity. When there were automatic looms, the mind was like an automatic loom; and, since young people in the loom period liked novels, it was the cheap novel that was degrading our minds. When there were telephone exchanges, the mind was like a telephone exchange, and, in the same period, since the nickelodeon reigned, moving pictures were making us dumb. When mainframe computers arrived and television was what kids liked, the mind was like a mainframe and television was the engine of our idiocy. Some machine is always showing us Mind; some entertainment derived from the machine is always showing us Non-Mind.”—Adam Gopnik, “How the Internet Gets Inside Us” (via marathonpacks)
Last night I went to a folk music concert in North Oakland. The walls were hung with woven tapestries and we all sat on futons. The room was packed was aging hippies: white ponytails and leather fringe. Aging hippies are a dime a dozen in the Berkeley environs. Heck, I was there with my aunt and uncle, real-live aging hippie leftists, who’ve been in Berkeley for decades. And so it wasn’t weird for them that almost every time the singer picked up his guitar, he’d allude back to the ’60s. He made it seem like that was when things were happening, not only in Berkeley, but across the country: protests and rallies and sit-ins and strikes.
It’s easy to think that younger generations are too complacent to share in the revolutionary spirit that existed in the ’60s. Maybe young people don’t care deeply enough about issues, or maybe we’re soft in the heads from too many years in front of the TV. Maybe we’ve passed the buck off to the Middle East because we’re too lazy to fight for rights anymore.
But maybe the problem is that we’ve grown up in a media culture that makes it seem like we don’t care and chooses to perpetuate that narrative. Maybe the media portray revolutionary aspirations in this day and age as pathetic or ineffectual, or worse: nonexistent. This is the conclusion I’ve come to after trying to read up about the Wisconsin protests. Many “reputable” news sources, like the New York Times aren’t highlighting coverage about the protests, either on their home pages or even buried deeper in the “National” sections of their sites.
I think the protests in Wisconsin are worth talking about, if not for the issues the workers are fighting for, then because they remind Americans that we can still fight. Why does Bahrain get the front page? What about what’s happening here?
Many protesters remained at the capital overnight again. When I left at 10:30 there were still 1500-2000 people watching the amendment hearing in the rotunda and protesting outside of the chamber doors.
The Amendment to Special Session Senate Bill 11 was passed at midnight last night along party lines (as expected), even after hours of scrutiny by democratic legislators.
Noam Chomsky says “Wisconsin protests might be beginning of a pro-democracy uprising in the U.S.” (via Democracy Now!)
I’m confused: there is a significant workers’ rights protest taking place in Madison, Wisconsin, where Governor Walker is trying to take away collective bargaining rights from teachers, EMTs, and other public servants. It appears that the New York Times doesn’t think this is a big deal; I can’t find a mention of it on their site. I kind of think it is a big deal, and definitely worth reading about.
This concludes this week's edition of Neuroscience Tuesday
I am all out of philosophical and scientific thoughts for the week. Perhaps I’ll devote tomorrow to other issues of great import and interest. Like, where does my DIY haircut fall on the spectrum of bad hairdos from Blagojevich to Bieber to Bardem? Stay tuned for gratuitous pictures.
I agree that language learning will someday be computationally tractable by machines…but for now, all we’ve got is Watson. You could even say that reading emotional cues that co-occur with language is a solvable computational problem (but not yet, of course). The study Lo talks about is Saffran, Aslin & Newport (1996) paper - it’s gone viral, by academic standards, with thousands of citations.
Lo, kids understand syntax too. Researchers created artificial languages to teach 12-month-old infants grammatical rules (Gomez & Gerken, 1999). For instance, they would play sentences like “PEL TAM JIC RUD TAM RUD” or “PEL TAM PEL JIC” over and over, until essentially kids learned grammatical rules by statistical regularities - e.g. “PEL” and “TAM,” when presented together, always take the same grammatical form and perhaps PEL is an article and TAM is a noun. In test trials, children were able to pick out valid and invalid NEW phrases with the nonsense words.
A postdoc in my lab is doing work on how kids can also infer the meaning of a word based on its grammatical context. For instance, 3-4 year olds kids infer that if a novel verb takes an object, it’s transitive, and if it doesn’t, it’s intransitive.
The really exciting thing is that so many things can be described by statistical models. For instance, my data so far in one of my studies shows that a classic econometrics model does a very good job of describing how kids understand other people’s preferences. Yet we all have to be wary of accepting these well-fitting models as the actual model being implemented in the human brain. This is why people are also wary of accepting Watson as truly ‘intelligent’ - he can beat us in Jeopardy, sure, but is he really doing things the same way we are? Definitely not. For one, we think in parallel, and Watson can’t. You could claim the 2800 computers running at once could simulate a small part of the human brain’s functioning, but unless we’ve made great strides in the year since my neuroscience class, we really don’t know enough about the brain to create anything but a very rudimentary computer model that mimics human thinking.
So when we talk about computational models, we need neuroscience to tell us whether our models are at all implementable or not. If anyone is familiar with Marr’s levels (1980), we need to figure out the implementational level of human cognition. We currently know quite a bit about algorithms that pretty well describe cognitive phenomena, but are those the very same algorithms our brain actually uses to solve the same problems, or are those algorithms just what we can feed into Watson to make him seem smart?
And even weirder, at what point does it stop mattering? (SKYNET.)
P.S. Screw you, computer, for not knowing words like “implementational.” THESE WORDS ARE IMPORTANT TO ME HOW COULD YOU
My dad once asked me whether I thought that the ability to learn language was special or uniquely human. I told him no. Obviously, I have a deep passion for language. I think it’s the coolest thing on Earth, giving our species unending delight and whimsy. But I don’t think the ability to learn language is a special skill. I’d argue (and I do, below) that the skills that allow babies to learn language are the same as those that allow a computer to recognize an object or a automated voice software to parse human speech.
I don’t mean to conflate the ability acquire and understand language with the ability to interpret the emotional content associated with language and discourse. My point is that I don’t think there’s anything about language acquisition that can’t be explained by computational learning. Yes, learning language is a huge feat for a baby. However, babies aren’t doing much else. They don’t need to worry about feeding themselves or keeping themselves warm. They don’t even need to move! The movements they do need are almost all reflexes. Babies aren’t worried about taxes, or their cholesterol or the score of the Sharks game. This means, that for the first few months of its life, the baby’s only job is to perceive communication. All of their mental energies can be focused on the task of acquiring language.
I believe that computational learning is the underlying mechanism for language acquisition.The acquisition of phonology (speech sounds and the rules governing speech sounds) is a great example. Through exposure to spoken language, an infant becomes attuned to sounds (phonemes) of its home language. It does this by selectively listening to the sounds that occur most often in its environment. At birth babies are sensitive to phonemic contrasts of all languages. By 9 months, the infant can’t perceive phonemic differences of foreign languages, and has become a specialist in differentiating the sounds of its own language.
The infant then applies computations probabilities to find boundaries in the continuous speech stream. Sounds that are more likely to go together indicate that those sounds go together within a word. Unlikely combinations indicate words on two sides of a boundary. When boundaries can be identified the speech stream can be parsed into discrete units, what adults know as words. Once the child recognizes words, it can use the same kind of computational learning as Watson to attach semantic knowledge to the lexical items. And once you have a few words in your semantic repertoire, they can be combined in ways using rules of syntax. I assume that kids put words together in computationally likely ways, but I don’t have data to back that up (reaganing: help a sister out).
Like Watson the Jeopardy-playing computer, humans learn by experiencing hundreds of exemplars and formulate rules that describe a likely prototype. If you see enough dogs, you learn that dogs tend to have four legs and fur and tails. What’s interesting is that we still see canines that have had a limb amputated, hairless chihuahuas or docked Dobermans as “dogs” despite the fact that they don’t precisely fit the prototype. This is because they share enough features with the set of things that are dogs to be included as members.
As a set of objects, dogs are variable. But speech is way more variable. Speech has been notoriously difficult for computers to parse because each speaker has a slightly different voice. There are differences in loudness, pitch, intonation, accent. The incredible variability of the speech stream is hard for computers to interpret, but toddlers do it pretty much effortlessly. It doesn’t matter if a speaker has a southern accent, or a cold, is a man or a woman or a child with a lisp: a listener hears past those surface differences and processes the stable core of the meaning of the spoken language. Just like we perceive all those outliers to be dogs, we perceive a huge variability of speech sounds as belonging to relatively few meaningful categories (phonemes).
The NOVA special about Watson explains that computers have only been able to master the speech recognition problem by using computer learning. Hundreds of speech samples of a given word are taught to the computer, which formulates a sort of “prototype” of that word: what are the features of the utterance that tend to occur in most of the exemplars. There may also be sets of features which co-occur. The computer induces these features and feature sets and when confronted with a new exemplar, tried to match it to an existing prototype.
Yes, Anonymous-Question-Asker from the depths of the ether, she is a vegetarian! Although not a terribly picky one. Like, that carrot that is drenched in buffalo sauce on the side of your hot wings is totally fair game, as are those noodles in your questionably vegetarian soup. Also, I eat the hell out of milk, cheese, honey, eggs and other animal-derived goodies, but nothing that was itself a critter. Sometimes people (usually Italian grandmothers) have difficulty grasping this fundamental tenet of vegetarianism. After I turn away a dish, they ask, “What about chicken?” No, I don’t eat any animals at all. “Well about fish?” Nope, not fish either; nothing that was a living animal. “But you eat mussels. Those aren’t really animals.” No…
I started being a vegetarian (over ten years ago, now) because I’d met a hot straightedge vegan boy. But I stayed a vegetarian because it’s a sensible and more sustainable life choice.
For the record, I post many vegetarian (and sometimes vegan) recipes on my food blog (whatwehadfordinner.tumblr.com). It’s a submission-friendly blog, so you can always send recipes in, veggie or not and I’ll post them up!
Romance and magic: Friday afternoon, Berkeley, California.
I had noticed the guy talking to himself, over by the full-bloom geranium planter. He was gesticulating liberally and seemed to be enjoying both sides of the animated conversation he was having with himself. He certainly wasn’t threatening: not on a busy sunny sidewalk on a Friday in downtown Berkeley. This man was obviously crazy, but refreshingly so. Not angry-crazy like the guy that road raged on me the other night on MLK; not creepy like the guy who walked past me downtown and - rather too loudly to be under his breath - whispered, “Her pussy was way too hairy.” And not disturbingly crazy like the woman back in Boulder with the dead eyes who drags around the rolley suitcase and talks gibberish to her shadow. This guy by the planter was pleasantly nuts in a way that allowed him to blend into the backdrop of the city street, to become just one more off-kilter being roaming Berkeley’s streets. I didn’t notice he had left his sidewalk post until he stepped out of the pharmacy. The automated doors swung to let him out and I saw that he was holding a single, long-stemmed red rose. He walked up too close and handed me the cellophane-wrapped flower.
"I pulled it out of the dumpster" he mumbled, and walked off shaking his head across Shattuck.
Other things I didn't know about whorehouses: You gotta know
My enquiries started several years ago when I spotted a New York Times article about a police sting on a Queens brothel. The headline caught my eye and I kept reading long enough to see that this brothel was run by an infamous Latina midget. I was appalled. How could I be living so close to such wacky-ass shit and have no idea? How could I be so naive? There were Mexican midget madams running whores only miles from my cozy suburban upbringing, and here I was, clueless. I didn’t realize there any were brothels in these parts, let alone ones like you might read about in some peyote-infused Kerouac ramble. Intending to never be thusly caught off-guard again, I vowed to learn where to find a hooker in my city.
I asked around and it turned out that everyone in town - apart from me - knew where to find a hooker. “Down by the Lincoln Tunnel- you can pick up a girl in the city, drive across to Jersey and then drop her off over there.” Or sometimes a terse, “The Dolphin Hotel on the West Side”. I got recommendations for myriad neighborhoods, nationalities and kinks. I learned where to solicit the best gay hookers and where I could have all sorts of predilections satisfied. Stupefied, I’d ask my informant (usually someone I’d previously assumed was a total square), “But how do you know this?” And the answer would come back, “You gotta know.”
Whorehouse words wholly unrelated: brothel and bordel
My mother was born and raised in France but didn’t speak much French to me growing up. Apparently my sister and I did our best to enforce assimilation into English-speaking culture by responding in French with put-on terrible American accents until she just gave up trying to speak to us. I do recall isolated French snippets, most of which I later discovered were vividly salty expressions or just plain curses. My mother was attempting to fake the neighbors out by yelling at us in French. They probably thought it was very classy, or something. One expression I remember is “Qu’est ce c’est ce bordel”. I understood this to mean, “What a mess!” but the literal translation of bordel is “whorehouse” (you’ve probably run across the Italian variant, bordello).
The other day, while thinking about whorehouses, I realized I’d assumed that brothel and bordel might share an etymological root. I mean, that “d” to “th” seems plausible as a historical sound change (see changes like Latin pater —> father; or German bruder —> brother).
But, like so much of my knowledge about whorehouses, I was woefully ill-informed. Brothel comes from Old English broþel, meaning “ruined or degenerate” or a person with those traits.Bordel is also an old word, but comes from Old French meaning a “cabin, hut or a whorehouse”.
According to the OED, confusion between the two arises because brothel originally referred to a person and bordel to a place. Combinations like “brothel’s house” and “bordel-house” ran together to form “brothel-house” (meaning whorehouse), which was shortened to the brothel we all know and love.