Psycholinguistics James Myers May 28, 2004 Brain and language OVERVIEW: 1. Brain anatomy 2. Neurolinguistic methodologies 3. Lateralization 4. The localization of neurolinguistic processes ============================================================= 1. Brain anatomy 1.1 First, some fundamental terminology: Neurology: the science and medicine of the brain (related to NEURON = brain cell). Neuroscience: just the science part of neurology, plus (sometimes) the study of artificial neural networks (i.e. connectionism). Neuropsychology: a branch of neurology that deals with the connections between the brain and behavior, usually using cognitive psychological models Neurolinguistics: a branch of neuropsychology that deals with language. 1.2 The basic fact that allows neurology to exist: The brain is not made of pudding. (That is, it has structure.) In fact, the brain has so many parts that you can't remember them all. Here are the major parts: 1.2.1 Cerebellum: the "little brain" near the back which has long been thought to be involved only in boring things like motor coordination. Like many other old assumptions about the brain, though, this may turn out to be wrong.... 1.2.2 Cerebrum: the big, famous part of the brain. The smartest part is the surface cortex which is only 1.5 to 5 mm thick! This is why the brain is wrinkled: it's so that this 830 cm^2 cortex can fit into your head! 1.3 More about the cerebrum: 1.3.1 The cerebral cortex is divided into two hemispheres that are connected only in the middle, in a structure called the corpus callosum. The hemispheres are connected to the body contralaterally: the left hemisphere controls motor movements on the right side of the body, receives auditory input from the right ear, receives visual input from the right half of the visual field (actually the left half of the retina of each eye because of optical inversion). 1.3.2 More about the cortex: Not everything happens in this cortical system; there are also subcortical connections: neural pathways that lead directly from one part of the cortex to another (like a "secret passage"). Damasio and Damasio (1992) suggest that this distinction accounts for Pinker's distinction between "connectionist" irregular inflection (cortical) and "rule-like" regular inflection (subcortical). More on Pinker and brains below. The cortex itself is structured: different parts of the cortex have different kinds of cell patterns (cytoarchitectonic fields or Brodmann's areas, which are often just known by their code numbers). Within the cortex, information seems to be processed in a parallel distributed fashion, as in a connectionist network (at least within each of Brodmann's areas). 1.3.3 More about wrinkles: Real neuroscientists have to memorize all those Brodmann's numbers, but lazy people like us can just use the big wrinkles as landmarks. First, some terminology: Gyrus (pl. gyri): a ridge (=long hill) on the cortex. Sulcus (pl. sulci): a groove (valley) on the cortex. Fissure: a deep and prominent sulcus. (Note: "fissure" is also an ordinary English word for any long, deep crack.) Now for the major landmarks: Central sulcus (the big vertical groove) Sylvian sulcus/fissure (separates the "wing") Angular gyrus: bends around the tip of the Sylvian fissure 1.3.4 These sulci and gyri sort of divide up the brain into four lobes (the liver and lungs also have lobes): Frontal lobe: from the front, back to the central sulcus Temporal lobe: the lobe under your temple, i.e. the "wing" that is separated by the Sylvian fissure. Occipital lobe: the "butt" of the cerebrum ("occipital" is medical Latin for "back of the head"). Parietal lobe: all the rest, between the frontal and occipital lobes ("parietal" is medical Latin for "wall of a hollow part"). 1.4 Finally, some (nonlinguistic) functional landmarks on the cortex: Sensory strip: just posterior -- i.e. towards the back -- of the central sulcus; the parts that receive skin sensations are arranged in a distorted map of the whole body. Motor strip: just anterior -- i.e. towards the front -- of the central sulcus; this also has a distorted map of the whole body, with parts for the mouth near the bottom (right above area 44). Primary auditory cortex: areas 41/42, i.e. near the (opposite) ear, at the top of the temporal lobe) Visual cortex: at the tip of the occipital lobe (far away from the eyes!) 2. Neurolinguistic methodologies Here are some of the major sources of data about brain function; other methods are also used. 2.1 Aphasiology: the study of aphasia: language disorders caused by brain damage. Step one: Wait until somebody gets brain damage from a stroke or an accident Step two: Determine the location of the lesion(s): the damaged part(s). (In the old days, this step required that you wait until the patient died to examine the brain.) Step three: Give the patient tests to see how the language abilities have been affected (requires that you know some linguistics, which most neurologists do not know) Advantages: It's been done for a very long time, so the procedures are very well understood, e.g. classic logic of double dissocation. Disadvantages: You may not know how "normal" the patient was before the brain damage. Subjects in a given study often vary quite a lot in the specific location of lesions; also lesions usually cover a much larger area than the one(s) you're interested in studying. Relative rarity and variability of aphasia often means that you must rely on case studies with small samples. 2.2 Electroencephalography (EEG): 2.2.1 Electrical sensors are placed on the skull in a specific pattern; they measure the electrical activity of neurons "firing". Advantages: Noninvasive and inexpensive Can give you very rapid measurements (one technique can make as many as 250 measurements per second); this allows for more flexibility with experimental stimuli. Disadvantages: Measurement is indirect, and so you can get false information from echoes, transduction along the skin, muscle movements, etc. 2.2.2 One special use of EEGs involves "evoked potentials" or "event-related potentials" (ERPs): the change in an EEG related to a particular stimulus, factored out from the background brain activity. The ERP can be used as an on-line measure of "surprise" caused by ungrammatical sentences, for example. 2.3 Neuroimaging techniques: 2.3.1 Computerized axial tomography (CAT or CT) scans: A regular X-ray machine that is connected to a computer; it rotates around your skull, so the computer can calculate an image of one "slice". Advantages: It's very common (there are about 10,000 machines in US hospitals), and it's useful for showing the location of brain lesions. Disadvantages: Like any X-ray machine, it doesn't show brain activity. 2.3.2 Positron emission tomography (PET) scans: The scanner measures radioactive particles injected into the bloodstream, and so can map out blood flow. Blood flow in the brain is actively changed over time, so blood flow shows which parts of the brain are most active; that's the biggest advantage. Disadvantages: The subjects must be injected with radioactive water just before performing the task, so you have to pay them a lot; also, risky for female subjects who may be pregnant but not know it. The radioactive water must be generated by a nearby cyclotron, since the half-life of the stuff is only about two minutes; thus the test is very expensive. Blood flow and radioactivity patterns don't change that quickly, so new conditions in an experiment can only be given every 10-15 minutes. This means that subjects cannot be given mixed stimulus list; must have 10-15 minutes of one sort of material, and then of another sort; i.e. can't make "movies" of people thinking. 2.3.3 Magnetic resonance imaging (MRI) scans: The scanner contains very powerful electromagnets that cause the protons in water molecules to "line up"; then the magnet turns off for an instant and the protons "flip back" and release a photon as energy: then their position can be detected by the photons shooting out of your head. Advantages: No radiation, not even X-ray-type radiation. Disadvantages: It's not good at showing function, so it's often used merely as a substitute for CAT scans, i.e. to map out locations of damage. 2.3.4 Functional MRI (fMRI) scans: Works basically the same as an MRI, except that it tracks where your blood has lost oxygen feeding hungry brain cells, so it can show activity like a PET scan but it's much faster. Advantages: No radiation or expensive cyclotron. It can capture images in as little as 30 msec! Disadvantages: Like regular MRI's, it's still expensive: one machine can cost several million US dollars, with yearly maintenance about US$250,000. You still have to wait for the brain to change patterns of bloodflow, so still no real "movies of thinking".... 2.3.5 Neuroimaging techniques have revolutionized the study of brain activity, and the main lesson that we've learned is: We don't know much. In fact, some things we were 100% sure about before has dissolved (see below). But some of this may be due to sloppy use of brain imaging -- it's so easy to collect data now that some researchers forget that they're supposed to be testing processing models. (One notable exception is Levelt & Indefrey, 2000, which describes a meta-analysis of brain-imaging studies to test Levelt's model of word production.) 3. Hemispheric lateralization 3.1 New Age nonsense: The "left brain" is rational and logical, while the creative "right brain" "retrieves the cosmic wisdom of mysticism", representing "the mind of the Orient" that overly logical Westerners must learn to develop??? Uh, sorry, nope. "One must caution ... that the experimentally observed polarity in right-left cognitive styles is an idea in general with which it is very easy to run wild.... It is important to remember that the two hemispheres in the normal intact brain tend regularly to function closely together as a unit." Sperry (1984). 3.2 The two hemispheres are not "specialized" for distinct functions; neither "dominates" the other; but instead certain functions tend to be lateralized: processed more quickly or efficiently on one side or the other. So "common wisdom" says this (e.g. Bever 1980): Left Right ---- ----- temporal information spatial information analysis into parts perception of wholes LANGUAGE 3.3 Some ways to study lateralization: 3.3.1 Dichotic listening: different sounds presented to each ear. Because of the contralateral wiring, a right-ear advantage implies left-hemispheric lateralization. Warning: sound from the right headphone can reach the left ear through bone conduction; also, there are also ipsalateral pathways between the ears and the SAME side of the brain. 3.3.2 The same things can be done visually: have subjects fixate on a central point, then flash stimuli on the right or left too quickly for eyes to move. Images in the right visual field will go to the left sides of each retina which are then processed in the left hemisphere. 3.3.3 Evoked potentials can be used to record differences in hemispheric responses. In one method, you measure the ERP to speech stimuli to see which hemisphere shows the biggests response. In another method, the "interesting" stimuli (e.g. speech) is played with "beeps"; the size of the ERP to the beeps is inversely related to the amount of brain resources spent on processing the interesting stimuli. So the hemisphere that shows the smaller ERP to the beeps must be the side that's processing language. 3.3.4 Split-brain studies: To prevent epilepsy (random neural firings) from spreading from one hemisphere to another, the corpus callosum used to be cut, since there were no major changes in behavior. However, if the two hemispheres cannot communicate, information processed in one side (e.g. left side for language) cannot affect the other side (e.g. motor control of left hand). Sperry (1968) started using these patients to study lateralization. Show a picture to left visual field (i.e. right cerebral hemisphere), and split-brain patient cannot name it. The same phenomenon occurs with haptic (touch) stimuli: if a patient holds an unseen object in the left hand (controlled by right hemisphere), patient cannot name it. 3.3.5 One more clever trick (discussed in Poizner, Klima and Bellugi 1990): In a shadowing task, subjects must repeat speech while they hear it. Subjects perform worse if they simultaneously tap with their right hand than if they tap with their left hand! 3.4 Left-hemisphere advantage for language seems to provide evidence that language is a module: a stimulus is only processed better by the left hemisphere if used as linguistic information: 3.4.1 Wood (1975): English-speaking subjects were presented with speech sounds. When the task was to judge the pitch (not linguistic), the ERPs to the stimuli were more active on the right. When the task was to identify the syllables (linguistic), the ERPs were more active on the left. 3.4.2 Klima, Bellugi, and Poizner (1988): brain-damaged ASL signers. Recall that ASL uses space to mark inflectional morphology: e.g. "she" (for your mom) might involve pointing towards imaginary point A, while "she" (for your sister) might involve pointing towards point B. Brain damage on the left affects the ability to use these inflections, but has NO effect on spatial cognition in general (e.g. can still draw maps). Brain damage on the right affects general spatial cognition, but amazingly has NO effect on the ability to use space as ASL inflection! 3.5 However, humans are not the only ones with "language" in the left hemisphere: Japanese macaque monkeys perceive vocalizations of their own species better via the right ear than via the left (i.e. the left hemisphere is better) [Petersen, Beecher, Zoloth, Moody & Stebbins 1978] This is even true for species only very distantly related to us, e.g. chaffinches (songbirds) [Nottebohm 1970]. 3.6 Why are communication systems lateralized to the left? 3.6.1 The more "modularist" answer: Maybe it's for the same reason that a hand has five fingers (and not four or six): i.e. "just because." Reptiles, birds, whales, etc all have "five fingers" due to some ancient accident of history, so maybe that's the same reason why "language is in the left". 3.6.2 The more "interactionist" answer: Maybe it's because of the general differences in brain lateralization noted above: Bever and Chiarello (1974): musically trained listeners (who may listen to music analytically) show a right-ear advantage for music perception; musically naive listeners (who may listen to music as "wholes") do not. While there is a strong right-ear advantage for consonant perception, there isn't for vowel perception (Shankweiler and Studdert-Kennedy 1967), maybe because vowels are longer (and so don't require temporal analysis the way consonants do?) Thus since most linguistic processing is analytical, that's why the left hemisphere is better at it. 3.7 OK, so which hemisphere is better at processing phonology, syntax, lexical semantics and discourse? 3.7.1 Phonology: more in left hemisphere. Split-brain studies show that the right hemisphere "knows" what words like "ache" and "lake" mean, but doesn't "know" that they rhyme (Levy 1974). Speech production is very difficult for the right hemisphere (as shown by cases of left-hemisphere brain damage), but this may be because the left hemisphere is better at controlling sequences of motor acts in general (Corballis 1980). 3.7.2 Syntax: more in left hemisphere. Split-brain patients were presented with pictures to the left visual field will not be able to choose which of the following pairs of sentences describe them [Gazzaniga and Hillyard 1971]: The boy kisses the girl / The girl kisses the boy The girl is drinking / The girl will drink 3.7.3 Discourse and pragmatics: more in right hemisphere. Kaplan, Brownell, Jacobs and Gardner (1990): right-brain-damaged patients have no trouble judging the literal truth value of statements, but do have trouble identifying the use of "literal falsehoods," like friendly joking or sarcasm. 3.7.4 Lexical semantics: in both hemispheres, sort of. Split-brain patients have no trouble understanding nouns presented just to the left ear (Gazzaniga and Hillyard 1971); this is consistent with the findings presented in Damasio et al. (1996), which imply that object names are processed in both hemispheres (using data from brain-damaged patients, plus neuroimaging studies with normal subjects). However, verbs are different: split-brain patients cannot respond to one-word verbal commands presented to the left ear (e.g. "Smile") (Gazzaniga and Hillyard 1971). Is verb processing harder for the right hemisphere because it involves syntax? More on this later. Both hemispheres may store lexical information, but they behave differently in practice: Burgess and Simpson (1988) Normal subjects were presented with words flashed either in left or right visual field. Primed lexical decision task: primes were ambiguous; targets were either related to dominant meaning of prime, or to a secondary meaning. Interstimulus interval (ISI) was varied to see which of meanings of the primes remained active. Results for right visual field (i.e. left hemisphere): The dominant meaning was strongest, and secondary meaning faded quickly (we already knew this). Results for left visual field (i.e. right hemisphere): The secondary meaning did NOT fade, but in fact got stronger over time! Conclusions: while both hemispheres have spreading activation, only the left hemisphere has "controlled suppression" of inappropriate meanings. [Does this involve those hard-wired subcortical connections...?] 3.8 Variation in lateralization patterns. 3.8.1 The effect of handedness: The left hemisphere is dominant for language in 97% of right-handers; but for left-handers, only 19% are right-hemisphere-dominant, 68% left-hemisphere-dominant, 13% both (Corballis 1991). Even if right-handed subjects have left-handed brothers or sisters, they process language in subtly different ways from other right-handers (Bever, Carrithers, Cowart and Townsend 1989). 3.8.2 Male-female differences in lateralization: Sex seems to affect brain lateralization, but not as much as handedness (Caplan 1987:351): right-handed women have more bilateral processing of language than right-handed men. For example, when men and women read aloud, men show left-brain activation only, while women show both sides active (Shaywitz 1996). This difference seems to be genetic, and does not arise from life experience. Shucard and Shucard (1990): 3- and 6-month-old infants heard speech with beeps mixed in. Measured ERPs to beeps as inverse measure of hemispheric processing of the speech (i.e. the larger the ERP to the beeps, the more processing must have been shifted to the speech). Results: infant boys showed a higher response by the right hemisphere to the beeps, whereas infant girls showed a higher response by the left hemisphere to the beeps. That is, boys were processing the speech with the left hemisphere, while girls were processing the speech more the right hemisphere. Implication: Even little tiny boys have language lateralized to the left hemisphere, whereas girls tend not to. 4. The localization of neurolinguistic processes OK, we got (most of) language in the left (sometimes), but can we be more specific about the location of language processing in the brain? 4.1 Are different words processed in specific places? Yes, sort of. For example, Damasio et al. (1996): Aphasiological part of study involved patients with deficits in concept-form mediation, i.e. they had trouble on a picture naming task but no trouble showing that they understood the concept represented by the pictures. Some patients only had trouble naming people, some only had trouble naming animals, and some only had trouble naming tools. They also found patients with deficits in more than one area, e.g. people and animals, animals and tools, or all three (people, animals and tools). The crucial finding is that they found NO patients who had trouble just with people and tools. This behavioral pattern correlated with their brain damage: patients with people-naming problems had damage in the tip of the temporal lobe; patients with animal-naming problems have damage in middle of the temporal lobe; patients with tool-naming problems have damage at the point where the temporal lobe joins the parietal lobe. Thus there are no patients just with people+tool problems because the "people" and "tools" areas of the temporal lobe don't touch. Note: These areas were not where they stored phonological or semantic information for these words, but just where the phonological and semantic information was linked up. The results were confirmed with PET scans of normal subjects. When subjects named pictures of famous people, the tip of the temporal lobe became active (in both hemispheres); when naming animals, the mid-temporal region became active; when naming tools, it was mainly the other end of the temporal lobe. Are these innate modules? Not really. Maybe the "people" area is partly hard-wired, but the other areas can end up where they are by simple "spreading activation," which will put semantically similar words near each other: animals are similar to people, and tools are similar to animals.... 4.2 Broca's area: the side of frontal lobe, just above tip of temporal lobe and just below motor strip (near areas for the motor control of jaws, lip and tongue). 4.2.1 Damage to this area causes Broca's aphasia, where speech is effortful and not fluent: "Lower Falls... Maine... Paper. Four hundred tons a day! And ah... sulphur machines, and ah... wood... Two weeks and eight hours.... working ... working... working! Yes, and ah ... sulphur. Sulphur and... Ah wood. Ah ... handling!" (Goodglass 1973) Some features of Broca's aphasia: Articulatory problems. But this may just because Broca's area is so close to motor control areas. Agrammatism: usually defined as having problems with grammatical morphemes and function words; more generally, agrammatic aphasics have difficulty with speech, and produce short sentences. Note: grammatical morphemes are also messed up in writing, showing that the problem is not purely articulatory. 4.4.2 So what does Broca's area do? Broca's aphasia as "expressive aphasia"? That is, is Broca's area for language production only? Evidence: Broca's aphasics seemed to have no trouble comprehending sentences like "The book that the girl is reading is yellow." Problem: When sentences are given that require more than just real-world knowledge, where grammatical morphemes are crucial for understanding the meaning, they have much more trouble understanding. Thus they cannot distinguish between pragmatically reversible sentences (Caramazza and Zurif 1976) in a picture-identification task: The horse that is kicking the bear is brown. The horse that the bear is kicking is brown. Is Broca's aphasia literally agrammatism (i.e. the loss of the rules of grammar: syntax, morphology and phonology)? Pinker (1991) notes that Broca's aphasics have more trouble reading aloud regularly inflected forms than irregularly inflected forms, controlled for frequency and pronounceability. This implies that Broca's aphasics have little trouble accessing the lexicon, but they do have trouble applying grammatical rules. Problem 1: All kinds of aphasic patients have lexical access problems (Elman et al. 1996). Problem 2: In spite of their difficulty in using syntax to comprehend, Broca's aphasics can make certain grammaticality judgments (Linebarger, Schwartz, and Saffran 1983): *I want you will go to the store now. *How many did you see birds in the park? Problem 3: Broca's aphasics show cross-linguistic differences, which suggest they retain aspects of their native language grammars (Bates 1991) 4.2.3 More on Broca's area and inflection: an example of the danger of trusting neuroimaging too much: Jaeger, Lockwood, Kemmerer, Van Valin, Murphy and Khalak (1996): a PET study of regular and irregular inflection in English Nine male right-handed native speakers of English Stimuli: a list of 46 irregular verb stems; a list of 46 regular verb stems; the lists were matched for frequency. NOTE: the items could NOT be mixed, since the PET scan had to examine separately the response to each list: all regulars, vs. all irregulars. Task: must say the correct past tense form. Some results (seem to support Pinker's views about inflection): Both tasks activated Broca's area (processes abstract grammatical marker "PAST TENSE"?) Irregulars activated about twice as much space on the cortex as regulars (the lexicon is "everywhere"?) Criticism (related to use of PET scans): Processing 46 regulars in a row is a lot easier than processing 46 irregulars, which all have different patterns; maybe this task factor explains the difference in brain behavior. If so, the results give no support to Pinker's views! 4.3 Wernicke's area: just below angular gyrus (upper back of temporal lobe, near auditory cortex) 4.3.1 Normally, Broca's and Wernicke's areas work together: they are connected by a subcortical bundle of nerves called the arcuate fasciculus (medical Latin for "arc-shaped little bundle"). [Damage to this connector is causes "conduction aphasia", which disturbs the ability to repeat what one hears and to speak spontaneously] 4.3.2 Damage to Wernicke's area causes Wernicke's aphasia: sort of the complement of Broca's: speech is fluent but makes no sense (including made-up words): "Boy, I'm sweating, I'm awful nervous, you know, once in a while I get caught up, I can't mention the tarripoi, a month ago, quite a little, I've done a lot well, I impose a lot, while, on the other hand, you know what I mean, I have to run around, look it over, trebbin and all that sort of stuff." (Gardner 1974) 4.3.3 What does Wernicke's area do? Is Wernicke's area just "receptive aphasia"? That is, does it involve comprehension and not production? Evidence: Whereas Broca's aphasics can often guess what you mean, Wernicke's aphasics mix up lexical semantics, so if you say "Touch your knee" they might touch their ankle. But they make the same sorts of errors in their productions, messing up semantics or getting the phonological form wrong: "table" -> "chair", "clip" -> "plick", "paper" -> "piece of handkerchief, pauper, hand pepper, piece of hand paper". (cited in Pinker 1994) Is it involved with lexical access? That would explain jargon aphasia: the making up of new words when you can't find the real word you want. The speculation of Pinker (1994:311): "Wernicke's area seems to have a role in looking up words and funneling them to other areas, notably Broca's, that assemble or parse them syntactically. Wernicke's aphasia, perhaps, is the product of an intact Broca's area madly churning out phrases without the intended message and intended words that Wernicke's area normally supplies." 4.4 Another fact to make us more confused: There is a double dissociation between nouns and verbs (Chen and Bates 1994). Broca's aphasics have more trouble with verbs: Show a cartoon of a bunny who is crying, and the patient might say: "Bunny ... tears" [keep nouns, drop verbs] Wernicke's have more trouble with nouns: This kind of patient might say: "That thing here, whatever it is called, it's crying." An obvious explanation: Broca's area processes grammar (e.g. syntax), while Wernicke's area is crucial for lexical access. Verbs are syntactically more complex than nouns (e.g. "see" vs. "look" -- remember Gleitman 1990?) while nouns are lexically more complex than verbs (e.g. "cat" and "dog" differ much more in meaning than do "see" and "look"). Problems with the obvious explanation: Broca's aphasics don't really lose "grammar" (above) Chinese aphasics show the same double dissociation at the morpheme (word-internal) level (where syntax does not apply) (Bates, Chen, Tzeng, Li, and Opie 1991; Chen and Bates 1994): Chinese Broca's aphasics made more mistakes with the verb morpheme in VN compounds, even if the whole word was a noun, while Wernicke's made more mistakes with the noun morpheme (except when the whole word was a noun): e.g. VN that is a noun: ­¸¾÷ e.g. VN that is a verb: ¸õ»R Why does this happen? Not clear. But it's not obvious how syntax could be the explanation.... 4.5 OK, so what are we supposed to write in our notes about Broca's and Wernicke's areas?? The admission of a modularist (Pinker 1994:311): "But to be honest, no one really knows what either Broca's area or Wernicke's area is for." The comments of some interactionists: Elman et al. (1996:383): "In fact, the number of so-called language-specific areas are multiplying on almost a daily basis. Every new functional imaging study seems to bring another language area to our attention.... This all leads to the conclusion that domains like language do not live within well-defined borders, at birth or at any other point in development." REFERENCES Bates, E. (1991). Special issue: Cross-linguistic studies of aphasia. Brain and Language, 41(2). Bates, E., Chen, S., Tzeng, O., Li, P., & Opie, M. (1991). The noun-verb problem in Chinese aphasia. Brain and Language, 41, 203-233. Bever, T. G. (1980). Broca and Lashley were right: Cerebral dominance is an accident of growth. In D. Caplan (Ed.) Biological studies of mental processes (pp. 186-230). MIT Press. Bever, T. G., & Chiarello, R. J. (1974). Cerebral dominance in musicians and nonmusicians. Science, 185, 537-539. Bever, T. G., Carrithers, C., Cowart, W., & Townsend, D. J. (1989). Language processing and familial handedness. In A. M. Galaburda (Ed.) From reading to neurons. MIT Press. Burgess, C., & Simpson, G. B. (1988). Cerebral hemispheric mechanisms in the retrieval of ambiguous word meanings. Brain and Language, 33, 86-103. Caplan, D. (1987). Neurolinguistics and linguistic aphasiology. Cambridge University Press. Caramazza, A., & Zurif, E. B. (1976). Dissociation of algorithmic and heuristic processes in language comprehension: Evidence from aphasia. Brain and Language, 3, 572-582. Chen, S., & Bates, E. (1994). The dissociation between nouns and verbs in Broca's and Wernicke's aphasia: Findings from Chinese. Technical Report 9404. Center for Research in Language, UCSD. Printed in Aphasiology, special issue on Chinese aphasia. Corballis, M. C. (1980). Laterality and myth. American Psychologist, 35, 284-295. Corballis, M. (1991). The lopsided ape. Oxford University Press. Damasio, A., & Damasio, H. (1992). Brain and language. Scientific American, 267 (September), 88-95. Damasio, H., Grabowski, T. J., Tranel, D., Hichwa, R. D., & Damasio, A. R. (1996). A neural basis for lexical retrieval. Nature, 380 (April), 499-505. Elman, J. L., Bates, E. A., Johnson, M. H., Karmiloff-Smith, A., Parisi, D., & Plunkett, K. (1996). Rethinking innateness: A connectionist perspective on development. MIT Press. Gardner, H. (1974). The shattered mind. Vintage. Gazzaniga, M. S., & Hillyard, S. A. (1971). Language and speech capacity of the right hemisphere. Neuropsychologia, 9, 273-280. Goodglass, H. (1973). Studies on the grammar of aphasics. In H. Goodglass & S. E. Blumstein (Eds.) Psycholingusitics and Aphasia. Johns Hopkins University Press. Levelt, W. J. M., & Indefrey, P. (2000). The speaking mind/brain: Where do spoken words come from? In A. Marantz, Y. Miyashita, & W. O'Neil (Eds.) Image, language, brain: Papers from the First Mind Articulation Project Symposium (pp. 77-93). MIT Press. Jaeger, J. J., Lockwood, A. H., Kemmerer, D. L., Van Valin, R. D., Murphy, B. W., and Khalak, H. G. (1996). A positron emission tomographic study of regular and irregular verb morphology in English. Language, 72, 451-497. Kaplan, J. A., Brownell, H. H., Jacobs, J. R. and Gardner, H. (1990). The effects of right hemisphere damage on the pragmatic interpretation of conversational remarks. Brian and Language, 38, 315-333. Klima, E. S., Bellugi, U., & Poizner, H. (1988). The neurolinguistic substrate for sign language. In M. Hyman and C. N. Li (Eds.) Language, speech and mind: Studies in honor of Victoria A. Fromkin (pp. 138-152). Routledge. Levy, J. (1974). Cerebral asymmetries as manifested in split-brain man. In M. Kinsbourne and W. L. Smith (Eds.) Hemispheric disconnection and cerebral gunction. Charles C. Thomas. Linebarger, M., Schwartz, M. F., & Saffran, E. M. (1983). Sensitivity to grammatical structure in so-called agrammatic aphasics. Cognition, 13, 361-392. Nottebohm, F. (1970). Ontogeny of bird song. Science, 167, 950-956. Petersen, M. R., Beecher, M. D., Zoloth, S. R., Moody, D. B., & Stebbins, W. C. (1978). Neural lateralization of species-specific vocalizations by Japanese macaques (Macaca fuscata). Science, 202, 324-327. Pinker, S. (1991). Rules of language. Science, 253, 530-535. Pinker, S. (1994). The language instinct. William Morrow. Poizner, H., Klima, E. S., & Bellugi, U. (1990). What the hands reveal about the brain. MIT Press. Shankweiler, D., & Studdert-Kennedy, M. (1967). Identification of consonants and vowels presented to left and right ears. Quarterly Journal of Experimental Psychology, 19, 59-63. Shaywitz, S. E. (1996). Dyslexia." Scientific American, 275(5) (November), 98-104. Shucard, J. L., & Shucard, D. W. (1990). Auditory evoked potentials and hand preference in 6-month-old infants: possible gender-related differences in cerebral organization. Developmental Psychology, 26, 923-930. Sperry, R. W. (1968). Hemisphere deconnection and unity in conscious awareness. American Psychologist, 23, 723-733. Sperry, R. W. (1984). Consciousness, personal identity and the divided brain. Neuropsychologia, 22 (6), 661-673. Wood, C. C. (1975). Auditory and phonetic levels of processing in speech perception: Neurophysiological and information-processing analyses. Journal of Experimental Psychology: Human Perception and Performance, 104, 3-20.