Higher Cortical Functions: Language (Section 4, Chapter 8) Neuroscience Online: An Electronic Textbook for the Neurosciences | Department of Neurobiology and Anatomy - The University of Texas Medical School at Houston

Chapter 8: Higher Cortical Functions: Language

Anthony Wright, Ph.D., Department of Neurobiology and Anatomy, McGovern Medical School

go back one page go forward one page

A major issue of the topic of higher cortical function is the degree to which mental and cognitive functions are located in specific brain regions.

We begin by looking at the neuroscience of language. Language is one of the most elaborate cognitive behaviors. The pioneer of brain localization of language function was a French neurologist, Pierre Paul Broca, working around the time of our civil war—1861. Broca extended a theory proposed by Gall (Franz Joseph Gall) that the cortex was divided into 35 or more regions for attributes such as hope, generosity, and secretiveness. According to Gall, these were cortical “organs.” They grew with use, much as muscles did with exercise, and caused bumps and ridges on the skull. Thus, was born the science of phrenology. Hundreds and thousands of skulls were collected and bumps were correlated to attributes and personalities of those individuals. From a study of these correlations, so-called experts would go out into the population and “read” the bumps of living people to divine their attributes and personalities—all for a fee, of course.

Broca’s work differed from Gall’s. Broca argued for examining brains of people with clinical disorders for lesions that might then indicate the brain region responsible for the disorder. Thus was born the field of neuropsychology which flourishes today and has provided much of what we know about localized function in the brain.

Figure 8.1
Above, the four lobes of the cerebral cortex. Click any label to see that area highlighted.

Figure 8.2
Cortical regions of some of the major language areas and pathways. Click any label to see that area highlighted.

In 1861, Broca described a patient named Leborgne who could understand language but could not speak. He had no motor deficits to account for his inability to speak. He could whistle, utter isolated words, and sing the lyrics of a melody. Postmortem examination of Leborgne’s brain showed a lesion in the posterior region of the frontal lobe. This region is now called Broca’s area. A few years later, in 1876, Karl Wernicke described another type of aphasia. This aphasia, or language disorder, involved a failure to comprehend language rather than a failure to speak. Broca’s patient could understand language but not speak, whereas Wernicke’s patient could speak but not understand language. Location of the lesion in Wernicke’s patient was different from Broca’s patient. It was in at the junction of the temporal, parietal and occipital lobes—now called Wernicke’s area. The junction of temporal, parietal, and occipital lobes is an important area to remember.

Wernicke proposed that language involves separate motor and sensory programs located in different cortical regions. The motor program, located in Broca’s area was suitably situated in front of the motor area that controls the mouth, tongue, and vocal cords. The sensory program, located in Wernicke’s area, was suitably surrounded by the posterior association cortex that integrates auditory, visual, and somatic sensations.

Wernicke’s model is still referred to today. According to this model, initial processing of spoken or written words takes place in the primary and unimodal sensory areas for vision and audition. This information is then conveyed to the angular gyrus of the posterior association area. This was thought to be the area where either written or spoken words were transformed into a common neural representation. Then, these representations were thought to be transferred to Wernicke’s area where they were recognized as language and associated with meaning. Without meaning, there can be no language comprehension. These neural representations along with their associated meanings are then passed along, via the arcuate fasiculus, to Broca’s area where it is transformed into a motor representation that allows for speech.

From the Wernicke model, Wernicke correctly predicted a third type of aphasia—conduction aphasia. Comprehension and speech articulation are spared but these two areas were thought to be disconnected. Patients can comprehend and speak, but they omit parts of words and substitute incorrect sounds. They are aware of their errors but are unable to correct them.

More recent results, such as those from functional imaging, have shown that the Wernicke model is an over simplification. There are, as we will see, other brain areas responsible for different aspects of language. We will briefly consider different aspects that constitute language before considering brain areas that control these different language aspects.

8.1 What is Language?

Language is communication through words or symbols for words. Words are an association between a sound and meaning. By 6 years of age children understand about 13,000 words and by the end of high school about 60,000 words. Thus, children connect a new sound to a meaning about every 90 waking minutes.

8.2 Is Language Innate or Learned?

Language is considered to develop in five phases.

1-4 months Cooing vowel sounds
5-10 months Babbling strings of consonant-vowel syllables (e.g., mamamama)
10-15 months First Words —consistent object labels
18-24 months Two Word Utterances meaningful pairs of words
>25 months Meaningful Speech

The issue is whether this development is innate or learned, and was most prominently raised as an issue by seminal books published in the late 1950s by Skinner and Chomsky. In support of innate development of language, all cultures learn language. Even children together in a social environment but deprived of any developed language—invent their own language. In support of learned language, young children deprived of language (e.g., parents are deaf or depraved) acquire language fully if learning takes place before puberty. If after puberty, they are very inept at language. If an infant has its whole left hemisphere removed, it will develop language fully in the right hemisphere. If adults have left hemisphere removed all language skills disappear forever. Finally, young children acquire several languages perfectly, whereas later in life it is much more difficult and with telltale signs of accents and grammatical errors. Thus, from the nurture standpoint there appears to be a critical period for language acquisition. From a neurological standpoint the developing brain is plastic with regard to language for several years.

Figure 8.3
Current source density measures showing activity flowing into the head (sinks) in purple and activity flowing out of the head (sources) in orange for grammatical violations of English phrases. Early English learners show highly left-lateralized activations, but later learners show increasingly greater bilateral activation. (Adapted from illustration in Neville, H.J. and Bavelier, D., Specificity and plasticity in neurocognitive development in humans. In: Gazzaniga, M. The New Cognitive Neurosciences, 2nd ed. MIT Press, 2000.)

8.3 When Did Language Evolve?

If language is innate then we can ask: When did it evolve? Language may have evolved as long ago as 2 million years ago. One of the things about language is that it is lateralized to an area called Broca’s area in the left hemisphere. Even the skulls of Homo habilis, a distant ancestor that lived about 3 million years ago, show an enlargement in the region of Broca’s area of the left hemisphere. If language evolved, should not a common evolutionary ancestor have passed this ability along to apes? Possibly, but on the other hand the division between apes and 6-8 million years ago may have allowed sufficient time for a separate development language in the evolutionary branch leading to man. One way to answer this question is to identify language in some other species?

8.4 Is Language Unique to Humans?

We can all think of language-like behavior from our pets. Dogs, for example, understand words like “come”, “dinner”, “fetch”, etc. But is this language?

Parrots have been taught to “talk.” One parrot, Alex, can name, request, or refuse more than 100 objects. When Alex is shown five keys or five corks and is asked verbally by his trainer, “How many?”, Alex answers verbally “five.” He can combine the concepts of number with color and can correctly answer the question, “How many blue key?” Alex knows the concept of same and different. When presented with a pair of objects that differ in matter (wood, plastic, metal), shape (triangle, square, etc.), or color (red, yellow, etc.) and asked “What’s different?” Alex will then “say” the correct answer. Alternatively, a blue wooden square and a red wooden triangle might be presented with the question “What’s same?” and Alex will answer “wood.” Alex can also be presented with novel shapes, colors, or matter and be highly accurate. But, is this language?

Finally, some of the great apes (gorillas, chimpanzees) have been taught symbols or tokens to represent objects and actions and others American Sign Language. These are hand signs used by deaf people, related to the hand signs used by the Native American Indian peoples. In either case they have learned the meaning of several hundred signs or symbols. Is this language?

The answer is probably not. They learn by rote behavior that if they want to get the treat (e.g., apple) that they have to put the prefix “Give me” before making the sign for “apple.” Apparently, they do not learn the meaning of the phrase “Give me.” What evidence would be necessary to show that they did learn the meaning? Related is that the word strings of sentences that these apes make do not increase with training as they do with young children.

This then brings up the question: If this is not language, then what aspect of language is missing? The answer is grammar.

Figure 8.4
fMRI brain images showing areas of increased blood flow when normal read English sentences, congenitally deaf ASL signers read English sentences, and congenitally deaf ASL signers view sentences in ASL. Hover over any subtitles to highlight the pathway. (Adapted from illustration in Neville, H.J. and Bavelier, D., Specificity and plasticity in neurocognitive development in humans. In: Gazzaniga, M. The New Cognitive Neurosciences, 2nd ed. MIT Press, 2000.)

8.5 Language Contains Grammar

If language is communication through words, then the next question is how are these words associated to provide this communication. The answer, of course, is grammar. Grammar is the rules that allow thoughts to be expressed in words and words connected into sentences when speaking and comprehending. Language is not how people ought to talk; it is how they do talk. For example, “ain’t” communicates as well as “isn’t.”

Grammar can be subdivided into morphology, phonology, prosody, and syntax.

8.6 Disorders of Language: The Aphasias

Aphasia is a language disorder. Such disorders revealed areas of the brain critical for language.

What makes language special in terms of associating its function with specific brain areas is that language is lateralized to the left hemisphere. If it were represented in both hemispheres then brain damage would be less likely to reveal aphasia because the other undamaged hemisphere would just take over language function.

Language is lateralized to the left hemisphere in about 96% of the population. This percentage is somewhat less for left-handed than right-handed people.

How was lateralization of language discovered? At least four lines of evidence have been important: 1) Neuropsychological studies of patients who suffered stroke, head injuries, herpes encephalitis, and degenerative diseases such as Alzheimer’s and Pick’s disease; 2) functional imaging studies of normal individuals; 3) electrical stimulation of patients undergoing neurosurgery; 4) split-brain patients.

Continued work on the neuropsychology of language disorders from patients with lesions and from stimulation studies of patients’ brains prior and during neurosurgery, has greatly elaborated on the Wernicke model of language. For example, Wernicke’s area has been subdivided into different functions. Lesions of the frontal-temporal area produce lexical deficits—problems in understanding meaning. Lesions of the parietal-temporal region produce syntactical deficits—problems in relating words in sentences.

8.7 Modern Framework

Implementation system: Broca’s and Wernicke’s areas along with areas of the insular cortex and basal ganglia analyze incoming speech in terms of phonemes and other grammar.

Mediational system: Areas of the temporal, parietal, and frontal association cortices that surround the implementation system. The mediational system fosters communication between the implementation system and conceptual system.

Conceptual system: Areas distributed throughout the association cortices that are important in learning, memory, and conceptual knowledge. These are associative areas considered in the next lecture, Higher Order Cortical Function: Association and Executive Areas.

Figure 8.5
Major gyri of the human cerebral hemisphere (O-opercular, Or-orbital, T-triangular, SPL-superior parietal lobule). Select any label to highlight the gyrus.

The different language and association areas are not as uniform in function as once thought. They have been subdivided, several times in some cases. It becomes awkward and imprecise to refer to these different brain areas in terms of lobes, sulci, and gyri. Many researchers use Brodmann’s numbering scheme. Korbinian Brodmann, a German anatomist, was inspired by Broca’s and Wernicke’s discoveries. At the turn of the century, Brodmann used cortical cytoarchitecture, the common appearance of neurons within the cortex, to identify 52 distinct areas.

Figure 8.6

Brodmann’s areas based upon the similar appearance of cells and cell layers of the cerebral cortex. For example, sensory areas have prominent granule cell layers (IV), whereas motor areas have prominent output layers (V). (see enlarged version)
(From Brodmann, K. Vergleichende Lokalisation lehre der Grosshirnrinde in ihren Prinzipien dargestellt auf Grund des Zellenbaues, Leipzig, 1909, J.A. Barth)

8.8 Broca Aphasia

Broca aphasia is damage to the inferior left frontal gyrus, which are Brodmann’s areas 44 and 45 including underlying white matter, insula and basal ganglia.

In many casual conversations, Broca aphasics appear normal in comprehension. Their speech, however, is obviously impaired. Speech is slow. Articulation is impaired and melodic intonation (Prosody) is lacking. The difficulty is not so much in pronunciation of individual sounds as in transitions from one sound to another sound. Thus, there is not a smooth articulation of flowing speech. In severe cases, perseveration of motor elements may be so great as to prevent pronunciation of simple words. For example, a Broca aphasic might say “2…2…2…2…want to…to…to…go…go…go…home.”

So far, this is according to the Wernicke model. But the normal comprehension is largely an illusion. Since most communication is with simple sentences, Broca aphasics make sense out of a sentence from only a few key words.

If Broca aphasics are given the sentence: “The apple that the boy is eating is red.” They can understand it. Boys eat apples. Apples do not eat boys. Apples are red. Boys are not red.

One can, however, show that comprehension is not normal by their inability to understand the sentence: “The boy that the girl is chasing is tall”. They have difficulty understanding this sentence because either boys or girls can be tall and because boys can chase girls or girls can chase boys. To understand the sentence you have to understand the phrase “that the girl is chasing” to identify whom is chasing and who is tall.

This is a grammar deficit, but grammar is by no means totally absent.

“John was finally kissed by Louise.”
“John was finally kissed Louise.”

The Broca aphasic is apparently able to detect that “John was finally kissed Louise” is ungrammatical because a passive verb (was kissed) is followed by a direct object.

Where they have problems is linking two elements in the sentence that refer to the same thing. For example, they cannot accurately discriminate between:

“The woman is outside, isn’t she?”
“The woman is outside, isn’t it?”

The actual problem that Broca aphasics have is thought to be a memory problem. Linking two elements in a sentence requires keeping the first element in memory. As the gap between elements to be linked becomes longer, these patients have more trouble.

Functional imaging results tend to support this interpretation that Broca aphasics have a short-term memory problem. A subregion of Broca’s area in normal subjects is more strongly activated (PET imaging) when the gap between elements to be linked is long rather than short. Some activity is apparently going on in this subregion to keep the elements in memory so that they can be linked.

8.9 Wernicke Aphasia

Wernicke aphasia is damage primarily to Brodmann’s area 22 at the junction of the temporal, parietal, and occipital lobes. This is the left auditory association cortex.

Speech is fluent, effortless, and melodic, unlike Broca aphasia. The content, however, is often jumbled. A Wernicke aphasic might say: “Aah dar, the spoon was needed for the telephone.” Even phonemes of individual words can become jumbled. In such cases the words are unintelligible—and are called neologisms. Such a patient might say: “Aoh confondo…noconfulo...he indent the confondo.” Wernicke patients have difficulty in word retrieval, and cannot select words that represent what they intend to say—this is called semantic paraphasia. They may say “headman” for “president.” They also have difficulty comprehending speech of others. It is useful to think of Wernicke’s area as a brain comprehension center, but as we saw with Broca aphasia there was a loss of comprehension there too.

Deaf people who sign with American Sign Language (ASL) use these same areas of the brain in the same way as those who speak language. Lesions in Wernicke area of deaf people loose ability to comprehend sign (ASL). Those with lesions in Broca’s area lose the ability to make signs.

8.10 Conduction Aphasia

Conduction aphasia is caused primarily by damage to Brodmann’s areas 39 and 40.

Speech production and comprehension are less affected than in Broca or Wernicke aphasias. They cannot repeat sentences accurately. And they have trouble naming pictures and objects. We will say more about these anomies in the next section.

Conduction aphasia is not produced by disconnection of the arcuate fasiculus alone as Wernicke thought. Damage producing conduction aphasia, however, does destroy many interconnections among the temporal, parietal, insular and frontal cortices that are responsible for assembling phonemes into words.

8.11 Transcortical Aphasia

Transcortical aphasia can manifest itself as transcortical motor aphasia and transcortical sensory aphasia. The name transcortical comes from damage to pathways in addition to damage of language areas. The pathways that are damaged are those that connect language areas to other parts of the brain.

A way to remember these transcortical aphasias is that the relation between transcortical motor aphasia and transcortical sensory aphasia is similar to the relation between Broca aphasia and Wernicke aphasia in terms of articulation vs. comprehension, respectively.

Transcortical motor aphasia is a disconnection of the language areas from areas that initiate and control speech. For example, these patients have trouble saying “kick” in response to “ball”, even though they use these same words correctly in ordinary speech. They can, however, repeat even very long sentences.

Transcortical motor aphasia is caused by damage to the left dorsolateral frontal area. This is an area anterior and superior to Broca’s area.

Transcortical sensory aphasics have fluent speech with impaired comprehension—very much like Wernicke aphasics. But unlike Wernicke aphasics, they can repeat sentences accurately.

Transcortical sensory aphasia is caused by damage to in the area of the junction of the temporal, parietal, and occipital lobes.

8.12 Global Aphasia

Like the term implies—it’s everything. Global aphasics have all of the disabilities of Broca, Wernicke, and Conduction combined. They cannot comprehend, repeat sentences, or speak meaningfully. They do, however, speak in expletives, and can recite such things as the days of the week.

Global aphasia is caused by damage to anterior language region, basal ganglia and insula (as in Broca aphasia), posterior language region (as in Wernicke aphasia) and superior temporal gyrus (as in conduction aphasia). Typically, this large amount of damage is caused by infarct in the middle cerebral artery.

8.13 Aphasia Screening

Below are some simulated examples of the capacity for spontaneous speech, comprehension, and repetition in aphasics.

Figure 8.8
Simulated examples of spoken language deficits in patients suffering with Broca, Wernicke, Conduction, and Global aphasia.
(Adapted from Kertesz, A. Western Aphasia Battery. San Antonio, 1982, Psychological Corporation).


Stimulus: "What kind of trouble are you having?" Stimulus:  "Please repeat this sentence:  The stray animal was timid."
BROCA "Well, um . . .  see . . .  I um . . . not sure." "Timid."
WERNICKE "I'm having some trouble." "Um ... eh ..."
CONDUCTION "I can't seem to, my sentences, the words having trouble saying."

"The dog was . . . what was that last word?"

("Let me repeat it:  The stray animal was timid.")

"The dog was ... um..."

GLOBAL (no words, only gestures) (no response)

View some video clips of actual aphasia patients:

8.14 Other Language Area

Dysarthria is the inability to control muscles of articulation and produces speech that is weak, flaccid and systematically distorted. Apraxia is a deficit in planning the desired speech movements and produces errors phonemically similar to the target word, for example saying “yawyer” for “lawyer” or “cookun” for “cushion”. Apraxia nearly always occurs in Broca’s aphasia and accounts for the slow effortful style characteristic of Broca’s aphasics. But recent evidence indicates that not all apraxia patients are Broca aphasics. There are, for example, patients who have apraxia but not Broca aphasia and these patients all have lesions of the insula (island of cortex) located at the superior tip of the precentral gyrus.

Figure 8.9
Percentage of overlap of lesions in 25 apraxia patients showing that all (100%) have insular lesions (yellow area). On the right are 19 control patients without apraxia. They have lesions surrounding the insula but none have lesions in the insula. (Adapted from illustration in Dronkers, N.F., Redfern, B.B., and Knight, R.T., The neural architecture of language disorders. In: Gazzaniga, M. The New Cognitive Neurosciences, 2nd ed. MIT Press, 2000.)

Specialization of language in the brain may be best shown by different areas of the brain that are active when naming different kinds of objects. For example, the naming of animals is apparently located in the medial occipital cortex whereas naming of tools appears to be in the left middle temporal gyrus.





Figure 8.10

A. Medial surface of left hemisphere showing activation in medial occipital lobe when subjects silently named animal drawings relative to tool drawings.

B. Medial surfaces of right and left hemispheres showing lesions in 28 patients with impaired recognition and naming of animal drawings

C. Lateral surface of left hemisphere showing active premotor and middle temporal gyrus areas when subjects silently named tool drawings relative to animal drawings.

D. Lateral surface of left hemisphere showing lesions of 8 patients with impaired recognition and naming of tool drawings.

The active left premotor site in C. apparently is due to subjects imagining that they are grasping or using a tool (e.g., a hammer). A study has shown that this area is active when subjects imagined grasping a tool shown in a picture with their right hand. (Adapted from illustration in Tranel, D., Damasio, H., and Damasio A.R. A neural basis for the retrieval of conceptual knowledge. Neuropsychologia 35(10):1319-1327, 1997.)


8.15 Right Hemisphere Involvement in Language


Some patients cannot place the appropriate stress, timing or intonation on words in their sentences. This condition is called prosody.

The right hemisphere controls prosody. It has an organization that roughly parallels the arrangement of Broca’s and Wernicke’s areas of the left hemisphere. The right inferior frontal gyrus is the site for articulating or producing prosody. The right posterior temporoparietal region is the site for comprehending prosody.

The sentence “Bill is here.” has stress on the first word, whereas the sentence “Bill is here?” has stress on the last word. A schoolteacher, for example, with a developing prosody deficit began to have difficulty controlling students in her class. She had feelings of anger when the students misbehaved and desire to assert her authority in order to reestablish control, but she could not convey these feelings because she could not place appropriate stress on her speech.

Some patients with prosody deficits cannot comprehend emotional tone or prosody of others that are speaking. These patients don’t understand jokes and have difficulty relating to people in social gatherings.

Alexia and Agraphia

Alexia is the inability to read and agraphia is the inability to write. These inabilities may be present individually or together and combined with aphasia. Vision is bilateral in the occipital lobes. Alexia results from disruption of visual input from both hemispheres to the language areas of the left hemisphere. PET functional imaging studies show that reading activates an area in the left hemisphere that is just anterior to the visual cortex. This may be the critical area that when damaged results in alexia.


Dyslexia is a reading disorder. It is often referred to as developmental dyslexia. Estimates of the population that it effects range from 10 to 30%. Its cause is not known for sure, and there may be multiple causes.

Figure 8.11
Asymmetry of the two lateral sulci showing the posterior extension of the left lateral sulcus indicating that the planum temporale is larger on the left. Select any label to highlight the area. (Adapted from Rubens A.B., Mahowald, M.W., Hutton, J.T. Asymmetry of the lateral (sylvian) fissures in man. Neurol. 26:620-624, 1976.)

One possibility is that, like alexia, there may be abnormalities in connections between visual and language areas.

Another possibility is that there might be a deficit in hemispheric specialization. In dyslexic males, unlike normal males, the left planum temporale is not larger that the right one. There is incomplete segregation of cell layers. There are clusters of neurons that appear to be inappropriately connected. Maybe the migration of neurons to the left temporal cortex during development was slowed or somehow disrupted.

Another possibility is that conduction velocity in the magnocellular pathway of the visual system is slowed. The cells in this pathway are abnormally small. Dyslexic patients have trouble processing fast, high-contrast stimuli, like words.




Test Your Knowledge

  • Question 1
  • A
  • B
  • C
  • D
  • E

Patients with difficulty selecting words representing what they intend to say (semantic paraphasia), along with difficulty in repeating sentences is most often associated with:

A. Broca aphasia

B. Conduction aphasia

C. Deficits in prosody

D. Wernicke aphasia

E. Transcortical aphasia

Patients with difficulty selecting words representing what they intend to say (semantic paraphasia), along with difficulty in repeating sentences is most often associated with:

A. Broca aphasia This answer is INCORRECT.

Broca aphasics have articulation problems, not word selection problems.

B. Conduction aphasia

C. Deficits in prosody

D. Wernicke aphasia

E. Transcortical aphasia

Patients with difficulty selecting words representing what they intend to say (semantic paraphasia), along with difficulty in repeating sentences is most often associated with:

A. Broca aphasia

B. Conduction aphasia This answer is INCORRECT.

While conduction aphasics have slight difficulties with word selection or repetition, they have many fewer than the correct answer.

C. Deficits in prosody

D. Wernicke aphasia

E. Transcortical aphasia

Patients with difficulty selecting words representing what they intend to say (semantic paraphasia), along with difficulty in repeating sentences is most often associated with:

A. Broca aphasia

B. Conduction aphasia

C. Deficits in prosody This answer is INCORRECT.

Prosody is not an issue here.

D. Wernicke aphasia

E. Transcortical aphasia

Patients with difficulty selecting words representing what they intend to say (semantic paraphasia), along with difficulty in repeating sentences is most often associated with:

A. Broca aphasia

B. Conduction aphasia

C. Deficits in prosody

D. Wernicke aphasia This answer is CORRECT!

Wernicke aphasics have many more problems with word selection or repetition than transcortical and conduction aphasics.

E. Transcortical aphasia

Patients with difficulty selecting words representing what they intend to say (semantic paraphasia), along with difficulty in repeating sentences is most often associated with:

A. Broca aphasia

B. Conduction aphasia

C. Deficits in prosody

D. Wernicke aphasia

E. Transcortical aphasia This answer is INCORRECT.

While transcortical aphasics have slight difficulty with word selection or repetition, they have many fewer than the correct answer.










Donate Now

Donations to Neuroscience Online will help fund development of new features and content.

go back one page go forward one page