Re: 1.) Explain PET...please. 2.) What happens to the brain at/around puberty?

Dear Jordi,
These are very complex and involved questions. I can say a lot about PET, but not much about critical periods.

The basic principle behind most PET imaging is the idea that when neurons become more active (i.e., they fire more action potentials) a greater amount of blood flows to the area where the active neurons are. By placing a radioactive tracer such as oxygen-15-labeled water into the blood, changes in blood flow in the brain can be visualized. So it is an indirect measure; it does not measure the actual activity of neurons, but blood flow changes. These blood flow changes have been shown by a great deal of animal and human research to be associated with increased regional (or "local") neuronal activity in the brain. The word "regional" is critical here. PET is a relatively sensitive technique. During basic sensory or motor tasks, PET can easily detect changes in primary cortical areas such as V1 for vision, A1 for hearing, S1 for touch, and M1 for movement. Blood flow to parts of the brain not involved in the task does not change. Of course, the technique has serious limitations. It has a resolution of approximately five millimeters or so. This means that two signals that originate in the brain half a centimeter or less from each other will appear to the PET camera to be one signal. There are thousands of neurons in a square millimeter of human cerebral cortex. So using PET, we can only map brain functions in broad outline.

PET and the study of language
There have been literally hundreds of PET studies involving language functions. The book Images of Mind, by Marcus Raichle and Michael Posner, has relatively easy-to-understand discussions of some of these experiments and covers a lot of other cognitive neuroscience as well, with an emphasis on functional imaging research.

A seminal paper was published in Nature in 1988 by Steve Petersen, Raichle, and others. This paper showed that different regions of the brain were activated when subjects saw words, heard them, spoke them aloud, or generated verbs that went with presented nouns. Over the past nine years, this last task, the "verb-generation" task, has been used by labs all over the world because it so reliably produces increased blood flow in an area of the left inferior frontal cortex.

Since that initial paper, there have been numerous studies of reading, listening, speaking, using a second language, looking at false fonts, looking at consonant strings, etc., etc. In a study by Ferruccio Fazio and others in Italy, people who had learned a second language after the age of seven showed different brain activations when listening to their native language compared to listening to their second language, even though their comprehension was roughly the same. Given the weight of all the evidence, I would say that PET is quite reliable for studying language processes.

Of course, this does not mean that using PET we can classify every single neuron in the brain as either being involved with a particular language function (or any other brain function) or not being involved. Far from it. But there is no doubt that functional imaging research techniques such as PET can contribute significantly to the scientific understanding of relatively complex mental functions such as language.

The "critical period"
Language acquisition may be the most-studied example of a critical period, but there are certainly many other critical periods for learning in humans. It is probably not a coincidence that most of the best classical musicians, tennis professionals, figure skaters, gymnasts, etc. started practicing their craft by age 5 or even before. Although I can't cite any research on this topic, I doubt that it's simply a matter of having had more years of practice. There seems to be something different about the brain up to age 7-12 or so that allows it to acquire all types of skills more rapidly and efficiently. And there is extensive research in animals as well as humans indicating that this is true for many lower-level brain functions. For example, depth perception is partially due to ocular disparity, the different location of a retinal image on the two eyes. If a person's eyes do not look in the same direction, their depth perception will be reduced. If the abnormality is corrected at an early age, the patient will have normal depth perception, but if it is not corrected until later, they will not.

As for whether the cerebellum plays a role, I really don't know. Until very recently the cerebellum was thought to be only involved in motor functions. But it is now fairly well accepted that it is also involved in cognitive processes, including language. Preliminary evidence suggests that the cerebellum may develop abnormally in autism. But the study of the non-motor functions of the cerebellum is in its infancy. Attempting to study the neurophysiological development of a uniquely human entity like language is extremely difficult because one is forced to use the most heavily protected research subject population there is: children. Fortunately, with the advent of functional magnetic resonance imaging, which is almost certainly a completely safe, harmless, and repeatable technique, even these types of questions can begin to be answered.

Petersen SE, Fox PT, Posner MI, Mintun MA, Raichle ME. Positron emission tomographic studies of the cortical anatomy of single word processing. Nature 331: 585-589, 1988.
Fazio F, Perani D, Dehaene D, Grassi F, Cohen L, Cappa SF, Dupoux E, Mehler J. Brain processing of native and foreign languages. Neuroimage 3: S583, 1996.

Joe Simpson
MD/PhD Student, Neurosciences
School of Medicine
St. Louis, Missouri, USA