wait, what did you just say?

Remember this sequence: 1 cherry 45 chair

If you remembered everything that ever happened, and perceived everything on this earth there was to perceive, you would be either or both: completely crazy, or a super human.

Your brain is very good at filtering out things. White noise (fans in the room, industrial lighting, your nagging in-laws), light, sensations (that pebble in your shoe you got 3 miles ago but weren't able to get out, the tag that itches), and even individual words.

Without looking, what was the very first word I wrote? What was the last sentence of that clause. Can't remember? Can you remember the context of that sentence?

This is known as working memory. It's kind of like the RAM in a computer (or, RAM is kind of like this, since we've been around much, much longer). It's a cache of random bits that you actively hold in your head. How does this apply to my Div3? You generally do not remember the words of a sentence, rather the sentence as a whole.

But what if that sentence doesn't make sense? Ah, here's where it gets complicated.

There is ample data that suggests those who can remember more items in their working memory (words, numbers, shapes, locations are all things you hold in your memory) are better comprehenders of language. They can untie linguistic knots such as: "The horse raced past the barn fell." Or "He was kind of spooked out so he went to the closet where his baseball equipment was. He saw a bat, it was brown and flying about the ceiling." I, of course, am copying this all from my own memory, so it probably doesn't matter.

What was the first sentence of that last paragraph?

Working memory has rapid, rapid decay. What was that sequence I asked you to remember at the top of this entry?

Oddly, there is weird evidence that suggests that the more you use it, the better you get. In my own experiments, the average college student figures out tricks to remember things more and more as we go along. But that's a whole other story.

There are three parts of the working memory system, as it stands. There have been variations and violations, but this is a theory that is generally well supported, and there has been neurological evidence that supports it, as well. One is the articulatory loop, the other the visuo-spatial sketchpad. These go into the central executive. The central executive does things like hold onto it, process it by retrieving things from other parts of memory (definitions of obscure words, whatever the last phrase might have said). The visuo-spatial sketch pad holds objects themselves in working memory: where you just put the keys down, the order of the piles of index cards in front of you).

1 cherry 45 chair


Working memory improves from childhood to adulthood. Look at babies' difficulty with object permanence (playing peak a boo is hilarious to a baby, for instance) compared with an adults complex knowledge of where everything in their life is located. This, of course, degrades after around age 21. But it peaks at 18, and plateaus until a graduale drop-off.

So in conclusion, working memory is essentially incredibly short-span memory. Short term memory is longer than WM. WM only holds about 6-7 items, and degrades as the brain's resources are drawn.

For further information, read:
Baddeley, AD (1986) Working Memory. Oxford Press, England.

More detail to follow, especially on development of working memory.

So, take it from the top...

Someone famous once said, "the best place to start is from the beginning."

The past entry kind of skipped that. So, if you're just tuning in, here it goes.

"Reading, The Brain, and Reading the Brain" is the title of my Div3 (aka thesis) project at Hampshire College in Amherst, Massachusetts. It combines three fields I'm intensely interested in: neuroscience/neuropsychology, language, and human development. I can't just tell you what it is, though. If I were to say to you, "It's an investigation into the link between working memory and the N400 in children and adults," I'd bet 10 dollars you wouldn't understand more than half of that statement.

Let's start with the basics: what is an N400? It's a component (aka, part of) an ERP. An ERP stands for Event Related Potential. An Event Related Potential is a specific electronic peak that your brain gives in response to various stimuli. That is, your brain reacts differently when it's analyzing a face versus when it reads a normal sentence versus is neutral. We've been able to track down various stimuli-response events, and they're temporally related, most times.

What is an ERP, though? Every moment of every day that you live (and sometimes, after death) your brain is sending electricity around itself and to your body. That electricity extends through your bones, guts, and skin. If we limit motion, we can isolate the brain's response to various stimuli. ERP's are measured in microvolts; using just your potential brain electricity, we'd need about 9,000 people to have the same charge as a AA battery. That's why millions of people charge what is known as the Matrix.

P=positive N=negative. P=N in terms of the effect. Just like negative current is the same
as positive. P100=P1=Positive charge 100 miliseconds post-stimulus. P3=positive wave 300 miliseconds post-stimulus. N4=negative wave 400ms p-s.

With me so far? So, to recap: brain gives off electricity, we can isolate electronic components (the waves are referred to as components because these timings aren't exact). The N400 is semantically related. That is, almost no other stimulus besides words makes the N400. Grammar doesn't even trigger the N400, that's been isolated to the P600, which is an up-and-coming field of research; that there could very well be hard-wired brain function related to grammar.

To learn more, read these:
Luck, SJ (2005) An Introduction to the Event-Related Potentials and Their Neural Origins
in An Introduction to the Event-Related Potential Technique. MIT Press, Boston. 1-50.

Kutas, M., Hillyard, SA., 1980 Reading Senseless Sentences: Brain Potentials Reflect Semantic Incongruity. Science 207(4427) 203-205

Osterhout, L, Nicol, J. (1999) On the Distinctiveness, Independence, and Course of the Brain Responses to Syntactic and Semantic Anomalies. Language and Cognitive Processes 14 (3) 283-317

I think that's enough for this lesson. When I'm back to procrastinating, I'll be back with you.

Mildly Academic Post

So, another come and go and so much for all that work I was going to do.

The thing is, I've found this amazing book that's a little more than half-relevant. The quote I posted before comes from it. Most research I can skim, this I just want to read every page of.

At the present, I'm confused about my study in that there are a bajillion ways I can go about it. I recently came back to the original material I used over the summer to form my hypothesis: that working memory in schizophrenics is strongly linked both neuro-and-psychologically irregular N4 patterns. Which opens up another door I don't have time or resources to walk through.

But it's compelling and I'm starting to see why my hypothesis holds water. If this study shows nothing, I want to do a developmental P6 study, since that hasn't been done yet at all so far as I can tell.

I also want to fund studies, but there is so little available to people who aren't PhD candidates. Basically, I want to achieve everything I possibly can ever.

This is my first academic entry

This made me really happy to read:
Regarding fMRI's: "It is assumed that the BOLD (blood oxygenation level dependent) response indirectly reflects neural activity. Although these methods represent major advances for the field of cognitive neuroscience, they are not without complications as tools for studying real-time language comprehension. First, the hemodynamic response to an event is delayed several seconds and eveloves over 10-15 s. Thus speakers produce (on average) three words, four syllables, and 12 phonemes per sec. Furthermore, the processing of asingle linguistic unit, for example, a word, most liekly involves a constellation of processes, each having temporal durations of considerably less than 1s [the time it takes for an fMRI to read neural activity in a human brain]" Osterhout, L., McLaughlin, J., Kim, A, et al. 2004 Sentences in the Brain: Event-Related Potentials as Real-Time Reflections of Sentence Comprehension and Language Learning in: Carreiras, M., and Clifton, C [eds.] The On-Line Study of Sentence Comprehension