Funded Grants

The functional neuroanatomy of reading: Comparative research across writing systems

Reading is at once both simple and rich--simple enough for cognitive research to have gained an increasingly clear picture of how it works; rich enough to yield important lingering questions to be addressed by the convergence of cognitive and neuroscience methods. One particular characteristic of reading can illustrate this simplicity and richness: It begins with a reader looking at marks that are encoded in a system--a writing system. The fact that the world has different writing systems is a fact that cannot be ignored if we want to achieve a full understanding of the reading processes. These marks, organized differently in different writing systems, orthographies, and scripts are processed by the brain so as to yield elementary units of reading, that, for a skilled reader, become readily combined in the comprehension of sentences and text. In this project, we use cognitive and neuroscience methods to shed light this process of turning visual marks into word reading.

Our central question is how does experience in reading matter for how reading works. Reading Chinese is a different experience from reading English. As an alphabetic system, English represents the basic sounds (phonemes) of the spoken language in its letters, and arranges these letters from left-to-right, each roughly representing a sound in sequence. Chinese, by contrast, does not represent these basic sounds. Instead a character represents the entire spoken syllable while it also corresponds to a meaningful word, in most cases. And the square shape of the character presents another contrast with the linear ordering of alphabetic letters.

Our project combines two research programs, one that has focused on cognitive studies of reading in Chinese and English with one that has focused on the functional neuroanatomy of word reading, as studied with English. The methods of the project are primarily the application of neuroimaging (fMRI) techniques to three populations of adult readers: Monolingual speakers of English; native Chinese speakers who speak English as a second language; and English speakers who are learning as adults to speak and read Chinese. The project aims to make focused comparisons among these groups in simple reading tasks to address basic questions about how the brain carries out word reading.

Neuroimaging research in English word reading has produced interesting results with some emerging consensus on brain regions that support reading. One particular result remains uncertain: The existence of a "visual word form" area in left posterior cortex (near the occipital-temporal boundary) that responds only to strings of letters that are words or a lot like words (pronounceable nonwords such as "klag"). Because of the different organization of the Chinese system, comparing Chinese reading and English reading may lead to clearer understanding of this word form area--whether its location and function are universal or dependent on experience with particular writing systems, for example.

The first set of experiments in our project will provide a series of critical comparisons concerning the word form area. Subjects from the three language background groups will view Chinese and English graphic stimuli that vary in their formal properties, especially the extent to which they are word-like-letter strings, pseudowords (pronounceable nonwords), and real words in English; false characters (real strokes but not real radicals), pseudocharacters (real radicals but in illegal positions), and real characters in Chinese. Of special interest is whether a word form area is found in Chinese that is specifically sensitive to both real characters and pseudocharacters, analogous to English or sensitive selectively to real characters only. The latter outcome would suggest that Chinese characters are perceived more globally rather than compositionally as seems to be case in English.

Equally interesting is whether learners of Chinese show a pattern more similar to English or to Chinese natives. Our learners will be in their first and second years of college level Chinese. We predict that they will be insensitive to the difference between real characters and pseudocharacters, because their experience has not been sufficient to support such a distinction. The monolingual English group should not differentiate any of these Chinese stimuli--all should appear as potential members of the category of Chinese characters. As for the Chinese language group, we are interested especially in whether their word form area for English and Chinese overlap, or whether they have developed two nonoverlapping systems. Second, is their English word form area sensitive to the same contrasts as English speakers or does it include only actual words? More generally, we also ask whether, when one learns a new system, it is separated in its functional neuroanatomy from the native language system.

Beyond the basic comparisons across these groups in perceiving word-like stimuli are studies that provide more information about the perception of words and other studies that examine other parts of the brain network that responds to printed words. In the first category are studies that will compare face perception, which also may be carried out within specialized areas within the object recognition system, with word perception. And, because written Chinese has pictographic origins-although very little pictographic content in its current inventory of characters-it is interesting to ask whether those characters that are more picture-like are processed by the same mechanisms that process abstract characters, for which there is no hint of pictographic content.

In the second category, studies that examine other parts of the brain's word processing network, are studies that examine phonological processes, the pronunciation component of words. These processes are pervasive in reading, even when reading is silent and directed at meaning, and even when the writing system seems designed to allow by-passing pronunciation, as Chinese does. In these studies, we ask about the role of expertise, both reading skill within one's native language and experience in a second language. In recent work with surface electrical recordings, we have found evidence that American students learning Chinese show differential brain responses to characters they have experience frequently in their curriculum compared with characters they have experience less frequently. In this project we want to test the hypothesis that this frequency effect reflects brain processes in both the visual form area and areas in left temporal-parietal and left frontal regions that support phonological processes. On one account, the left frontal region is associated with more effortful phonological decoding (of the kind an adult might do for an unfamiliar word) whereas the temporal-parietal area is associated with more automatic word-level phonology. We want to test the hypothesis that the learners of Chinese, in a pronunciation task, show activation in the frontal area for less familiar characters with activation in temporal-parietal areas emerging for frequently experienced areas.

Another approach to phonology will take advantage of the fact that Chinese children are taught reading initially through pin-yin, an alphabetic system using letters of the western (Roman) alphabet. As adults, although theses readers experience little pin yin, they are able to read it. For an Englishspeaking subject, a pin yin word can be an English nonword. For example, san would a Chinese word (minus tone) for a Chinese reader, but a nonword for an English reader. If Chinese adults see a word like san written in pin yin, will they process it in the same way they would an English nonword, or more like a Chinese character? We expect the monolingual English reader to process it like a nonword. In another use of pin yin, we will examine whether learners of Chinese come to treat English nonwords as more regular in pronunciation. The inconsistencies of English spelling allow a nonword like "hint" to be pronounced in a "regular" way (as in hint, lint, and tint) or in an exception way (pint). Given a list of nonwords like "bint," subjects can produce mixes of pronunciation (regular vs exception). Our interest is in what happens after they take the course in Chinese, which begins with heavy use of pin yin, which is completely consistent or regular in its mapping. It is possible that their approach to English word reading, and English nonword reading, will be modified by a sustained experience with a completely regular system. In addressing these questions we rely on behavioral measures (pronunciations of words and nonwords) with some additional neuroimaging work targeted for both left temporal-parietal and left frontal regions.

Although our main focus will be on differences in expertise across systems, we will also be testing readers of English who differ in reading skill. It would be of great interest to discover that skill differences within English, which we can asses with our behavioral measures as differences in literacy experiences that produce differential knowledge of word form and meaning, are similar to the kinds of differences that arise in language expertise. In simple form, the question is whether English readers learning Chinese and monolingual readers of English show similar brain correlates of reading performance.